Just another Temple Fox MIS site

Weekly Question: Week 5

Leave your response as a comment on this post by the beginning of class next week as well as your comments on your peer’s responsesRemember, you need to average four posts a week for a B. For these weekly questions, I’m mainly interested in your experiences and opinions, not so much particular “facts” from the class!

Answer one of the questions below:

  1. Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
  2. Where do you stand on the initial issue of the case: Is the development of Watson similar to most information systems projects, or is it something fundamentally different? Defend your answer.

91 Responses to Weekly Question: Week 5

  • (1) For Meaningful Use Stage 3 there is a requirement to make data accessible regardless of the source db. My team had a requirement from the medication reconciliation team to make the data from our documentation db accessible to them in order to report “reasons for not delivering specific medications for specific conditions” that were documented and saved to our db. Based on the very generic requirement the developers on my team met with their project manager and she identified to them the high level workflow and based on that there was an agreement to create a new stored procedure call that the medication reconciliation team would call and request the data and the code would return the data based on specific parameters we outlined. The team believed that this was going to be a very straightforward solution. As the our developers completed the code and I started unit testing I realized that there was a missing piece since the data requested was not appearing on the medication reconciliation report. Since I am the project leader across the teams I had to go back and gather the teams together and again walk through the workflow. This time I had the medication reconciliation analyst and tester who worked on the original workflow attend the meeting and show us their test scripts and design specifications in detail. As we went through the steps I was able to quickly identify possible breaks in the process. There was another team involved in creating the workflow. This team created a triggering rule that pulled the requested data from another db source. In order for the new workflow process to work correctly the team who developed the triggering rule needed to change their code to identify the data was coming from our db source and not the old.

  • Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?

    When I was accepted by Temple University, I got the package which contains the introduction of the processes that I need to fulfill before my MBA journey. The most unclear requirement is preparing the financial statement. Based on the document, international students need to offer the document which proves they or their family have the ability to pay for program tuition and living expenses. However, the package did not specific address the term of financial statements or other requirements in detail. Most of Asian have multiple bank and investment accounts. Should we provide the certificate of deposit or we can simply use other statement such as real estate registration certificate? What is the minimum amount of saving that we need to provide for visa application? Those are the questions we need to figure out.
    At first, I approach to the administration to clarify the requirement. Unfortunately, they were not sure about the answer. They offered the way that most students did in the past but revealed that the formats were different by countries.
    Then I reached to current international students in the same program and then did the research on social media channels related to “studying abroad”.
    Finally, I checked all the information and picked the format that most students from Taiwan took.
    The hardest part of the process is to explain the difference about financial systems and cultures between counties to the administration. I found it is better to provide and ask for examples while conducting the communication. I also provided my feedback to administration office so that they are able to share the information to students in the future.

    • It sounds like you did the best you could to find out what the process was and how to identify the financial requirements. I probably would have done the same thing and reached out to my fellow students. It’s unfortunate that the administration who created the rules how no means to help you. Seems a bit of a gap on their end if Temple is going to have an international student program. This should be a “Lessons Learned”, organizational process asset, that the University should look back on and fix.

    • Pei Yen, relying on examples was probably the best course of action. I think that applies in the working world as well, especially when faced with a new task. Looking at the completed version or portions of the process used for a similar assignment is helpful in figuring out how you want to approach the problem. I am interested in learning more about the international student application process for our group project.

    • I agree with Colleen – I am very interested in learning more about this process in our group project. It sounds like a tricky process that is unnecessarily made even more difficult, especially when you throw in cultural differences. Before speaking with you the other day about this topic I had no idea there were so many additional hoops for international students to jump through when applying to US schools. I am particularly interested in your mention of financial systems. I know financial systems vary from country to country but I am curious how much different the terminology and overall systems and practices are. It makes me think about ordinary financial transactions (ex: buying a house or car) and wonder how difficult it would be to do those abroad.

    • I can totally relate to Pei Yen’s point, as even I had faced a similar issue while finishing the financial eligibility requirements for Temple University. Even I had to exchange mails back and forth with the admission committee, and the international student services. I remember discussing this with my future classmates and I found out that even they were equally clueless. I asked a few of my friends who have been studying abroad which helped me a lot in understanding what needs to be done. I also relied on social media channels, blogs and discussion forums dedicated to students studying abroad. But most of them were not sure whether the same process applies to every university or if they have a different minimum amount of savings specifications.

    • 2. Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
      One example of a problem I faced due to unclear requirements was when I was working in the Customer Support team. I was responsible for resolving the technical issues encountered by the customers. The problem that I faced here was that there were no defined requirements which made figuring out the solution challenging. We used to receive several issues faced by the customers and they all needed different approaches to solve the problem based on what the basic reason of the problem was. Most of the time, the error reported to us is just that the tool is not working. Without knowing the exact problem, it was difficult to understand what exactly the issue could be.
      I can relate the approach I used to solve this problem to the concept discussed last week – 5 Whys. I used to ask questions till the reason of the issue is clearly identified. This approach really helped me find the solution. But, there was also a bit of guess work involved and I had to evaluate several possible criteria.

      • Hi Sadhana, thanks for sharing your experience with us. I have not work for customer service but have been a customer to had issues with something and called the help desk. I find what you say is true, because when I spoke to the rep I found myself answering a lot of questions before the rep offered me a solution. I never stopped to think that maybe there workflow sent my call to that rep and they were just trying to understand why I was calling them. Thanks

      • Hi Sadhana, a totally relatable problem as i was also involved in the maintenance and support work of the projects. When we used to get tickets from the clients, their definition of the problem was the symptoms caused by the bug in our system. I am not sure whether we should consider them as requirements or the source to generate requirements. I agree with the approach of the 5 whys, which always comes in handy and even unaware of this approach, we tend to use this to get to the cause of the issue. A similar case that came to us was the loss of some attached documents from the system which was caused by a symbol which was included unintentionally in the name of the document that was turning it into a javascript. After putting the javascript check on the text box and asking the user to upload the file again, the issue was resolved. But shall we consider the problem statements as the requirements, although i am with you that most of the time the problem statements are vague which leads to undefined requirements.

    • Hello Pei, thanks for sharing your experience with us. I totally agree with your comments. Communication lies at the heart of all problems. You would think the administration would have time to make corrections to the requirements if the recruited the help of the students that were already admitted to the program. Also, being different cultures and financial structures I can understand how difficult it is to satisfy all this requirements. It is interesting that at the end you are able to figure things out when they seem impossible at the beginning. Best

  • Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?

    One of the latest and greatest examples of a solution which had unclear requirements would be the Fox MC project I did last semester. I worked with the City of Philadelphia’s Department of Commerce, and helped them build a strategy around commercial revitalization of two corridors in Philadelphia. In multiple meetings with the client, and a few email exchanges, it wasn’t clear if the client wanted a guidebook, with strategies, or a macro-level strategy in general. Our team spent weeks figuring out what the exact deliverables were. For the project, we were expected to speak to tens of stakeholders from different arena who would directly and indirectly be affected by revitalization projects. The team spoke to community members, real estate developers, city officials, and a bunch of people. What was fascinating was that with every stakeholder interview we did, more information was unveiled and more there was better clarity. The team in one of the first few classes had created a resource breakdown structure and an organizational chart, which helped us figure which piece fit in where. The team was also able to attend a couple of meetings where a focus group of community members were present along with other stakeholders. I believe as we kept interviewing more project stakeholders, we were able to connect the dots and make sense of what the client really needed, rather than what they wanted. We kept documenting what various stakeholders believe what they wanted, were able to facilitate a few requirements workshop. At the end, we delivered a fantastic project report to the client, which showed the client where the gaps were, and how they could be filled in with the strategic recommendations that we provided. It was one of the most ambiguous projects I had worked on, but with a bunch of intelligent people, trying to pull chords from different sides, I was able to do justice to the project. A few weeks within the project, we were able to design a system, where we knew what information could be gathered from which type of source, and how that would fit into the puzzle that we were trying to make sense of.

    • Your example is a great representation of consulting in general. (Deloitte even mentions on its site that applicants need to be comfortable with ambiguity.) A client will request a somewhat vague or seemingly insurmountable deliverable and you need to figure out how to accomplish the goal within a reasonable time frame. I am glad you and your team were successful! Creating a resource chart sounds like a great way to combat the obscurity of the assignment.

    • I agree with Colleen. Nice example of what you deal with as a consultant. Consultants tend to get a high level overview of the project and you have to discover exactly what is going on once you meet the client and what your real responsibilities are going to be. I can see how the requirements gathering was probably difficult especially since you had so many stakeholders to deal with.

    • Hi Sid, I was really impressed while I saw your final presentation. I think your team nailed it and handled the situation well. Your experience also reminded me my EMC project, which we failed to manage our scope of work and unclear requirements in the early stage. I worked in the project for a national insurance company, which wants to understand overall behaviors of the front sales forces as well as the right segment of the force to sell the latest term product. Due to the unique and complex value chain of the industry, our client (policy provider) is not able to control or even get to know more about the salesmen (producers) who sell the policy to end user. Even we had the memo describing the overall pictures of the project, whenever we met or presented our progress to the client, they would make new requirements for us. In addition, the client refused our proposal to conduct interviews with sales before designing the close-end survey and insisted we should utilize their previous survey version for online survey. Finally, we figured out that the political issue in the client’s organization played a huge role during the project. Each project related stakeholder in the company tried to get personal benefit in the project. We still completed the project, however, most of the team members would agree that we could provide more profound insights if we were not trapped in trying to fulfill all the requests from clients.
      What I learned from the project is that clarifying unclear requirement is the crucial first step for project management. Additionally, how to integrate and manage the conflicts as well as individual expectation among stakeholders is also the key that leads to project success.

      • Hey Pei Yen, being on the same EMC project with you, I couldn’t agree more. What made the requirements unclear was that we got 3 different requirements from 3 different stakeholders. And they changed their mind every-time we met them. There were two stakeholders who were on the marketing side, and very data driven. They wanted to know how we arrived at the data, what our assumptions were before we even went ahead with the analysis. The third stakeholder was their boss, and was focused on the strategy aspect. Unfortunately we did not have many interactions with her to know what her expectations were. Few members of the team met her during the initial stage of the project. No one heard from her after that, she was listening in virtually during our interim and final presentation. I think if we had her input, we would have taken the project in a different direction.

    • I am going through our Management Consulting project for Marketing right now and my team had a very similar experience. We are working for the Temple Alumni Association, and at first were very unclear of the requirements. For a for-profit organization, the measurement of success is more straight forward, but what metrics does an alumni association use to measure success. Do they want to see more involvement? More donations? When we first met with our client we spent a lot of time discussing what success looks like for them and what deliverables they would like to see by the end of the project. Eventually we were able to narrow in on target segments that they would like to grow and learned what performance indicators they would like to see. By utilizing research and eliciting requirements from the client we were able to focus in on the scope of the assignment.

    • Sid, it sounds like your team did a great job and learned a lot in the process. I am currently doing my MC project and my team is working with a local CSA. At first the client’s scope was incredibly large. After doing some research we quickly realized the proposed scope was unfeasible with our time and resources. So, we had to figure out what was possible and narrow down the laundry list of questions we could answer. Once we did that we spoke with the client again about what we thought we could accomplish. Now we are at the next stage – figuring how to do that! It is certainly a learning process!

    • Haha Sid, EMC is one of the best example and a learning experience to figure out the exact requirements and to narrow down the scope of the project. We spent first half of the semester actually figuring out the requirements and what is needed by the clients, as most of the time even they are unsure about the requirements. Several meetings, visits, interviews and lot of research leads us to the final requirements after which we develop strategies to present them to clients and deliver them. Most of the time we see that the requirements keep on changing, which we have to cope up with.

  • I am currently involved in an enterprise-wide project whereby security systems (physical access to buildings/hospital units/rooms) are being standardized. A gigantic sub-project is the rebadging of all personnel which includes approximately 18,000 paid personnel and 14,000 unpaid personnel (contractors, vendors, students, visitors). An individual will be required to have a valid HR profile that will be interfaced to the new security system in order for a badge to be generated (and consequently applicable egress access permitted). I am charged with building a new “company” in our HR system (Lawson) to house bona-fide “vendors” (think: Otis Elevator, LF Driscoll construction etc) that currently are commingled with other non-paid personnel. Understandably, there are so many downstream and upstream impacts beyond the scope of this blog. It was evident from the start that I needed to stay abreast of the developments of the parent project as I built so I can quickly identify requirements. Like so many of the blogs from last week, our organization – and my own department at times – works so quickly without necessarily taking the time to assess, regroup, and rethink as Ferrucci did with Watson. I “manage up” well enough so that my manager willingly keeps my team in the loop and allows us to poke holes in perceived down/upstream processes. Sometimes the poking reveals an “a ha” moment; other times, not so much. This is how I am personally figuring out what are my requirements for the build – beyond the obvious! Not all organizations allow sufficient time and money for everything so I firmly believe it is my job to work as effectively and efficiently as possible and use my position, experience and reputation to get the answers.

    • Hi Maria,
      That does seem like a tough project, especially since vendors can have many differences between them, and since they aren’t employees I would imagine it would be hard to integrate them into all parts of the HR system since some parts would be things that are specific to employees only. I normally take a similar approach when faced with unclear requirements. I try to ask questions where ever I can and gather as much information as I can, even if it is unorganized or unclear I can get it down first and try organizing it gradually as I learn more.

    • Maria, that’s a great point that you stated. Its not everytime that a organization has time for everything or if there are issues and the business is getting impacted, time becomes a crucial factor. In such cases, your connections and the way you coordinate with all the parties involved is highly important. In my prior organization, there were scenarios where the applications were not accessible and business was getting impacted. In order to firefight the situation, I had to gather information from Operations team (which might be involved in their daily activities and other requests) and from the vendor, at the same time keep the application owner updated about the progress. Having good connections across various departments enabled me to provide the solution efficiently to the client.

  • While working on a consulting project, my team was tasked with analyzing the competitors in the consumer management space. The client even though shared information regarding their own offerings and platform, there was not much clarity about their system and the industry. The team started off with the analysis by researching different players in the market and got initial understanding of the consumer management industry. There were also interviews that were conducted with the people who had worked in that industry. But still it was hard to propose any recommendations as there was no proper understanding of client’s system. So the team connected with the client and asked to organize a session about the intricacies of their platform. One of the employees at client end, explained step by step different parts of their platform and the advantages that were there with the platform. After the session, the team could understand the strength and weakness of the client and also visualize their position and accordingly provide the recommendations.

    • Interviewing is a good tool and technique in project management. It’s a good way to get understanding of an application or industry like in your case.

  • From what I understand about system development the case of Watson is different. It is not water fall development, does not seem to fountain or spiral. It seemed to be more like parallel development. It is interesting that they also felt the need to bring in someone with no knowledge of their in house development and wouldn’t let him read up on their current work. They wanted to force him to construct a system from scratch. Also, in many cases information systems have set requirements on how to interpret data. One the unique things about Watson is that the system must be able to interpret data in a variety of ways. This requirement caused the designers to carefully consider which data was to be entered into Watson, and required them to design a series of complex algorithms to analyze the data sets. The Time line for Watson was also problematic. Traditional development styles would have taken far longer than the project allowed. By opting for parallel they were able to create different parts of the Watson system. When I read the article about Watson, it really made me thing that they were attempting to create a brain. It has some of the rudimentary functions of a brain, we all compare the question to our own knowledge (database) and asses the probability of answer based on that knowledge. The major difference is that we are better able to think in abstract terms, and understand rhetorical context.

    • Since I work in software development Watson’s development process was not completely uncommon. You’re right that it was not waterfall. Parallel development actually happens a lot and when that happens connecting the pieces can be difficult. The situation with James Fan was interesting since he was an expert in the field but was used as the outside comparison group and, in a sense, was allowed to build his own version of “Watson.” I think the team had more process than mentioned but the article made it seem like they were almost completely unstructured in their approach. I was the lead analyst on an NLP project a couple years ago and for something like that you have to have some level of structure but need to be extremely flexible.

    • Hi John,
      I also found it similar that the way Watson was designed to come up with answers is similar to how our brain works in some ways. I think that the potential applications for Watson are very interesting. I read that Watson can be used to assist physicians as a clinical decision support system. With AI becoming more and more advanced, some believe that within 50 – 100 years there will be a technological singularity where computers will become self aware. It sounds like science fiction to me, but it could be a real possibility!

    • I feel your point John, “computer as a brain” . Some human beings cannot even think abstractedly. We are not an encyclopedia or a dictionary. It is going to be very difficult to create a computer that’s a brain. A brain with so much data to process at the flick of a button. The human who is managing, supporting or troubleshooting this brain, will need to have a vast amount of knowledge and capabilities.
      There is always going to be a dead-end, where the computer brain cannot figure out the answer to. The jeopardy questions will be created in a more difficult manner, if a computer is the player. Think of “SIRI”. He tells me all the time, that he doesn’t understand.

      • Hi Susan and Jon,

        I actually think a true A.I. will arrive sooner that 50-100 years. What I see when reading about he development of Watson, is the very beginnings of mind creation. The groups at IBM have correctly determined that a mind is in part composed of knowledge, which Watson does not lack, and they have also correctly determined that Watson needs a series of thinking processes to determine answers to questions. I think they are quite a more processes involved in thinking, and personality. I also believe that the creation of an artificial mind is progressing down two very different pathways. Take for example the mapping of a mind (https://en.wikipedia.org/wiki/Mind_uploading), where scientists are trying to digitally upload minds. To me it seems like parallel development of A.I. True, one could argue that one mind is organic to digital, and the other is simply digital, but mapping a brain can certainly progress the creation of digital mind. More to the point, the definition of A.I. will probably change as development continues.

      • This is such an interesting question- whether a computer can equate to a human brain. That is more of a philosophical question (which I personally enjoy pondering- but I digress). The interesting thing about this case is how the IBM team used this disconnect to define the issue, and ultimately reach a conclusion. They saw that traditional programs of the time approached data inquiries in a systematic and structured way, and how that did not reflect the human experience (ie chess vs. jeopardy). Because they were able to see this disconnect, they were able to circumvent the historical constraints of computer programming to create a program that was more flexible and intuitive. It is this very philosophical question that seems to have been the backbone of their process.

    • Hi John, I agree with you in that building Watson reads like an artificial intelligence project. It made me realize of how comlex our brain really functions. I like the two guys in the garage model to think of the problem in a different way. Many times there are more than one solution to a problem. The time constraint was interesting because the have to think and act effectively and use their time wisely. Thanks

  • I believe that the development of Watson is fundamentally different from development of most information system projects for the following reasons.
    First, Watson is based on machine learning. It is meant to answer questions posed by a human which involves processing human language, weeding out the ambiguity and seeking answers to the “intended” meaning of the question. The challenge here is that Watson has to start with ambiguous questions. Most information systems are designed to output answers to unambiguous questions and are bounded by rules. Watson on the other hand has no rules, hence adding to the vagueness. Second, development of most information system projects follows the traditional software development cycle – requirements gathering, specification, design, implementation and testing. However the requirements in the case of Watson were unclear since different answers could lead to different solutions. Also traditional information systems have a defined set of database, hence one will get the same solution even if the query is designed differently. But for Watson, each approach of querying will lead to a different answer. Ferucci compares Watsons functioning to the way the human brain works. A tradition information system is more linear in that – an input will lead to a specific output.

    • Swetha,

      Well put! I also thought it was interesting that each data base seemed to be queried in different ways. Also, they were acutely aware that some data sets would prolong Watson’s response time or cause a drop in confidence. When they encountered these situations that would actively remove them. Though I would say, one of Watson’s biggest pitfalls is that the system has no process for storing experience. Take for example, the question that caused embarrassment. Watson’s will answer that question wrong every time, where as a Human should be able to answer it correctly going forward.

      • Actually Watson does have a process in place to learn from experience. As Swetha mentioned, Watson learns by utilizing machine learning. Doing so it is able to analyze examples of Jeopardy questions to come up with its own set of rules of how to interpret the different idioms and puns used in their questions. Although, in some case they would need to go in and change/ add rules when Watson was off track. If you have the time, watch this documentary, although long it was very interesting and described the process much better than I can.


    • Hi Swetha,

      I somehow tend to disagree with you on this. Although the way the system for Watson was design differently, it was based on similar set of rules. Watson was supposed to give answers that were specific, and not ambiguous. It was just that the number of rules and the way they were applied were different. Even if the question is ambiguous, Watson looks for different permutation and combinations and comes up with a specific answer. I agree that the software development cycle might be different for Watson, as it is more linear. But when it comes to your point of databases, Watson does use various databases (the internet, various papers, etc.) to look for answers. Also, data is continuously fed into Watson, for it to become more precise in answering questions; while rules are continuously being added/modified to help improve its confidence.

      • Hi Sid,
        I agree with you that Watson might have some loosen rules for decision making.
        What I am curious is the techniques how Watson examines and scores the hypothesis. How did Watson weight particular answer more than others? What criteria (rules?) or logic help Watson to make the judgement? Additionally, how did design team identify which “rules” can bring positive impact and which rules can’t?

        • Pei Yen, this might be a complete guess but I will give it a shot. Think Google Now! You ask Google a question; there are 3 things that happen. Google gives you a precise answer, Google gives you a ton of search results, Google says that there is nothing it has found. Now take the first scenario. When Google gives you a specific answer, I believe Google has created algorithm to give priority to specific websites because they are more reliable. If that isn’t the case, Google looks around the web for multiple answers. If there is a common pattern that Google finds, it believes that it could be the right answer. If you try asking Google Now questions, there are times when seldom it gets the incorrect answer. The way this algorithm is designed is it keeps multiple factors in mind- reliability, how interconnected the pages are, and how many clicks happen on that specific website. That is exactly how SEO works too. If your website’s pages are interconnected (not just within the website, but also externally), and more people click on your website, Google shows that on top in the search results. I believe IBM followed a similar principle while designing Watson. Just to lighten things up a bit, here is the link to a video of Watson showing its prowess over humans. 


    • I agree, my response was similar to yours. Even if you take into account other systems that use natural language processing, say Siri for example, it is bound by a straight forward question with a definite answer, and if a specific answer isn’t found you would get general information on that topic. In Jeopardy the “answer” is given and the question has to be stated. While this is the same concept, it’s much harder to decipher a Jeopardy clue since it is a vague statement under some type of category where the hints in the clue are things like figures of speech, wordplay, etc. things that it is very hard for a computer to make sense of, but that a person may pick up on quickly.

    • I agree Swetha, I also think Watson is different than other systems. The case study mentions that in traditional systems there are specific queries that have specific answers in return. Watson was not like this and was more towards the knowledge side. In the article, “Systems Development Cycle” it discloses that knowledge end workers would not benefit from systems such as the waterfall. It works with actions that are automated and need specific inputs for the answers. Watson was given various questions that will always vary from each other thus putting specifications on such a system would not work and was not what was occurring.

  • Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
    When I started my current job, I knew all my nursing language and diagnoses but I had no clue about the specific job that I was hired to do. I was supposed to get a 3 week orientation. My boss showed me the 4 different applications we had to use. It all went right over my head. He rounded in the hospital with me and I was just shadowing him, just watching and listening; trying to figure out the language he was speaking. I didn’t know what questions to ask. He spoke with me about the issues and what the planned interventions were but I could not comprehend because I have never done this. I was in a daze. I was unconsciously thinking, can I do this job with such high expectation, if I don’t even understand what is happening. Am I able to help turn a whole department around? After the first week, I was on my own as my boss was constantly busy. I slowly started going through the patient charts and started writing down everything I saw that I thought was important. That was basically everything. I had no clue that these charts were scanned every few days and that all the information in the chart and operative reports and so on; could be viewed by staff who had security access to that information. I did not know, who my sister departments were. I didn’t know how my work affected the entire organization. It would have been nice to have known all those things before I put my hands on a chart. I just did not have any down time to catch up on educational reading material. As the weeks and months went on; I started asking questions of everyone who I thought could help me learn my job. I asked my colleagues and other disciplines at the hospital. Here I am at my 18th month since I started this job and I have changed a lot of my processes of doing things. I would say that I only feel about 75% confident.

    • Susan, I think the feeling you are describing is familiar for many people starting new jobs. I had a similar feeling when I started my MBA program at Fox. My background is in journalism and creative writing. I had minimal experience with Excel. On one of the first days of orientation, we had a data analytics workshop where the instructor showed us PowerMap and other interesting components of Excel. I was still trying to figure out how to add an equation so I felt completely lost. Like you, I asked a lot of questions (and have continued to in each quantitative course since then) and now I am much more comfortable with Excel. The biggest hurdle I found was that “you don’t know what you don’t know.” Asking questions is imperative, but sometimes you don’t know where to start. I hope you get to 100% soon!

      • Hi Colleen, I can agree with all you said. My bachelor’s degree was in the sciences and now I am going for Health Informatics, thus a lot is new for me. I am used to how my science classes were, how they taught, and what they taught. I am new to this program overall and a lot of the topics that are discussed are not something that I have been exposed to before. I ask myself many questions too and find myself researching the answers to them. There is a lot to learn about many programs like Excel or Access, which was revealed to me last semester. I know the basics of these programs, but I am sure there is a lot I do not know in depth and am not even aware of what these programs can do, which was quite interesting to learn about last semester.

    • Susan,

      I can definitely relate to you. More than once while working at my present job I have been given extra duties with 0 training. I was given a rather complex report of aggregate immunization data for providers and told “You’ll figure it out.” I had to ask as many questions as I could, and pull from every available public resource I could get my hands on. I also tried to compliment that by going to every optional training session and off site activity, but I recall feeling something very similar to what you have described. I also have to agree with Colleen, one of the hardest things for me to figure out was, “what are the right questions to ask?”

      • Colleen and John
        I have been in Healthcare for 25 years now. There are always new initiatives every time we turn around but no one wants to put the cash up. This is why training is so in-efficient. There is not enough time and there is not enough staff to train. I feel as if, the key is with the financial team. No one thinks of the the end-product, the frustrations and the in-inefficiencies; until they have to do a root cause analysis for an unexpected event; as we have previously discussed.
        Hopefully with all of you, young entrepreneurs in this class; there will be a brighter future ahead.

    • It sounds like you did all the right things, especially with asking a lot of questions. I know it is incredibly overwhelming when you are thrown into a situation like that and barely know who to direct your questions to, let alone figure out what questions to ask first because you have so many. However, your example makes me wonder how much different your learning would have been with formal training. I know that formal training is often very helpful but on the other hand a lot of people seem to learn best while on the job. Certainly a balance between the two is nice but I wouldn’t be surprised if you picked processes and procedures up more quickly the way you were thrown into it – and because you had no choice!

      • It is hard to learn everything in such a demanding job with all sorts of requirements that one has to meet. I think that experience teaches one a lot. Even with my job, my training was for a couple of weeks, but even after that and still today I learn something new or I discover something that I had no idea about and had zero training for. I think learning on the job is one of the best ways to learn because even if one makes a mistake they will remember that, at least for me that was the case. Training is vital, however training cannot teach one every single detail of a job, especially in healthcare I assume because things are always changing.

  • Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?

    This weekend I was asked for advice on applying for MBA programs by my girlfriends brother. It reminded me of how when I first decided to go back to school for my MBA, I had no idea what the requirements were as far as the process to get into a program or which program was best suited for me. I didn’t know if I would be better off in a part time, executive, or full time program. At first I wasn’t even sure what test I needed to take. I started by consulting a “subject matter expert” my friend who had recently graduated from Villanova’s part time program. He was very influential in my decision to go into a full time program based on my interest in making a career transition. I also did several internet searches to find out the information that I needed for the GMAT. By filtering through information, I was able to get an idea of how the test worked and what would be a good target score for myself. I later refined my school search by gathering additional information, I narrowed down my list of potential schools based upon rankings and location. I also took several practice GMAT tests to familiarize myself with the process.

    • Great point about consulting a “subject matter expert.” I did the same thing when I started looking at MBA programs. I spoke with people about the process, the school they had gone to, what/if any GMAT prep course they had taken, etc. I also spoke with program faculty and admissions staff, as well as other potential students–any stakeholder I could find. I find that I am doing the same thing now while looking for internships. Speaking with people at different levels (entry level and manager as well as currently employed at the company, previously employed, rejected from employment) has given me a better view of the companies where I am applying.

    • When deciding to go to grad school I also had to do a good amount of research. The requirements were unclear because I wasn’t sure what I wanted to do or where I wanted to go. As I looked into it more and narrowed down my options the requirements became clearer. I eventually decided on Health Informatics and so far it’s given me a better perspective of the industry I work in.

      • I had a similar experience John. When I finished my undergrad degree, I wanted to go to graduate school for something different. What that exactly was I was unaware of too until I researched a lot. Even with the research like you said a lot of the requirements were unclear and even what one can do with a certain graduate degree were unclear when I was researching various programs. That was until I too discovered Health Informatics and made the same steps in researching and found it rather interesting with a lot of various career options.

    • I think its kind of funny to think about processes like the one you explained and how something that used to be so foreign is now second nature. This is especially true for me with the GMAT. I did countless online searches for tips and message boards… and that was after searching for the best prep programs and taking one. I think a lot of the time we underestimate how much we learn while figuring out something that is unclear to us.

      • I feel that all of you in the MBA program are awesome. It is a difficult program. Those of you who are full-timers; dedicate yourself to achieve the finest. Back when I started college, I did not know anything. I had no one to show me the way. Today there are i phones, internet, you tube and contacts who can lead you to the next destination. There is a lot of knowledge out there. We just have to make the time to capture it and put it to good use.

    • Hey Thomas, I can relate to you on figuring out the requirements for MBA schools. Back in India, there is a culture of going for coaching classes if you wanted to give any public exam. GMAT classes are a huge business there. They have packages for students starting with only coaching, coaching +B school consulting, coaching + B school consulting + application processing etc. There are also online forums that discuss everything right from recommended GMAT books to groups for students joining a particular school that year. It is basically a GMAT one stop shop at no monetary cost. When I decided to do my MBA I spoke to some of my friends and looked up the online forum. It pretty much told me everything I needed to know to get started. Next I searched for coaching centers that were close by and conducted weekend classes with decent faculty. A few weeks into the classes I felt it was not very beneficial since the weekend schedule did not work out. So I bought the materials, and enrolled in online courses. While giving my exam I started researching on universities. Again this was driven by internet search. Looking back, I think I understood the process best when I just jumped into it and figured out what worked for me by trial and error.

      • I had the similar experience with Swetha. In Taiwan, we have some institutions and websites that provide the consulting service for application to business school. They will hold seminars and introduce the full process from test preparation to school selection. However, I found students should also consider about the bias issues from these resources. For example, for business reasons, those institutions will promote the number of clients that got into TOP 10 MBA programs and in the seminar. They would tell you regardless of your interests, you should always consider top ranking school. They would also tell you that in order to get into top school, you need to at least get 700 in the GMAT. With that high score, you don’t need to worry about other requirements such as application essays. I figured out most of the information is not that correct after I talked to current students and started to get interview invitation. I think for MBA application, it is important to understand the requirement to complete the process. However, it is more important to find the school that can fulfill your requirement for your future career.

    • Hi Tom,
      I think we can all relate to figuring out the MBA application process. I went to an information session and had no intention of applying until next year since the application deadline had passed. They told me there were still seats available but I had to get all my information in ASAP. At that point I had not even started anything! I had to collect my documents, transcripts, schedule my GMAT and get my references in a matter of 2 days. I agree, this application process is tricky and made even worse on a very tight timeline!

  • Based on the methods in which Watson was developed it seems to be different that most information system projects. The traditional Waterfall model takes a more linear approach, where the output of one stage is the input for the next, but building Watson was a very complex project that required many components such as different types of algorithms and knowledge sources. The requirements were very open ended and difficult to define specifically. Watson can be considered a type of artificial intelligence because it can learn, process natural language, and use reasoning to answer questions. The way in which Watson takes an input of information and outputs its answers and the confidence in the answers reminded me of a software development method in itself, and it has to do this all within a matter of seconds.

    IBM used a very large amount of computing power to run massive parallelism, which allowed Watson to weigh different factors like location, passage support, classification, and reliability. It also used shallow and deep knowledge. A certain percentage of questions were straight forward enough that it could find the answer with high confidence by just searching its databases, but the ones that required taking the context of the clues into account had to use deep knowledge where it would build evidence in support of an answer. In my job, we use the Agile method for many projects. This method involves collaboration between teams and also needs to be flexible and adaptable which is similar to IBMs DeepQA strategy, however most projects in a work setting don’t have this level of complexity which is why I agree that developing Watson is different.

    • Well said Jonathan. I agree with your point that the way in which Watson was built was different so it is not feasible to use the traditional Software Development Life Cycle here. And also not all usual projects are so complex and require such different approach in dealing with it. Watson had to be trained to “think” from multiple perspectives. The requirements for Watson cannot be limited and it can’t be restricted by applying certain rules. Limiting the requirements would limit its ability to analyze the clue and provide the right answer.

  • Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
    When I was working at a media company, my boss quit suddenly with only a week’s notice. I was promoted to fill his position and found that all of his projects were a mess. This was particularly bad for a catalog project that had started months earlier and needed to print in two months (a short timeline considering that nothing had been created yet). The catalog was the largest we had ever produced. The client was disorganized and always missed deadlines. To make things worse, the publisher/president of my division instructed me not to tell the client we had lost information when my boss left so I had to piece together information from what little documentation I had. I looked at the contract, spoke with the graphic designer and met with the sales rep to gather more information on the project. I was able to piece together most of the information, though we ran into several roadblocks along the way. The catalog was far from our best, but the experience was a lesson in backing up all contracts and other pertinent project information.

    • Sometimes being “thrown into the fire” can be the best way to learn. I had a very similar experience where my boss took a sudden and unexpected leave of absence when he had a serious illness in his family. All of the sudden I was put in control of the entire operation at work. Not having a crutch to lean on, I had to figure out my own methods to get work done that I was unfamiliar with. I had to work a few rather long shifts since I was suddenly tasked with double the work load, but in the end it was a great learning experience for me. I also gained some more respect for the behind the scenes work that my boss had been doing.

    • Hi Colleen, I had a similar experience in the past. When I moved from my past job to my last one, before getting hired, I had a condition that I wanted to work on mobile products. Within a couple of months of new job, the company announced a mobile app that would be launched soon. My colleague who was supposed to work on the project, resigned on the same day of the announcement. Although I had little to no knowledge how mobile products were built, I volunteered out of curiosity. What I did not realize was I was getting myself into a system of multiple approvals and processes. I struggled for a month, asking people what data to fetch from where, and what approvals to get from whom. Although those were tough times, those were also one of the best learning times at my job. It took me long to figure how things worked in the organisation. I spent most of my time gathering requirements, and asking people the silliest of questions. What kept me going was my the end goal- to get the app up and running. It was a great experience as I not only interviewed stakeholders about their area of expertise, but I was also able to build upon my knowledge base. I finally ended up being one of the very few people in the organization who knew what fit where, and which stakeholder had control over what element of the mobile products. I think sometimes taking the plunge helps us discover the best in us.

      • Sid, I agree that sometimes the questions I was asking felt silly, but getting over that feeling was the only way to get the answers I needed. Part of the problem was being in a company culture that did not embrace questions. Clearly, since my boss refused to let me be honest with the client, I was not in a particularly open environment. It is great that you were able to turn your confusion into expertise and that you became a resource for the company.

    • Colleen, this certainly sounds like a learning experience! I’m sure that after going through that convoluted project you did everything you could to never re-live it with future projects. On the flip side, I’m sure when you left the position you left your successor in a much better position than what you were given when you took over!

    • That’s kind of a similar experience that I had sometime back. While working post acquisition of a client by my previous employer, one of the critical elements of the project during that phase was gathering the information for the applications that were in place and ensuring that there were minimal glitches and outages (since the new team did not had much information to support the systems). And in this period one of my colleagues had to leave the project due to personal reasons, so I was assigned with additional applications to handle. So I started to put in extra hours to venture into the new areas, coordinated with the vendors and stakeholders to capture the information. I created a knowledge base for these applications and outlined protocols for handling customer requests and implementing changes in the different environments. I was able to put in fixes to the applications such as the jobs in the server were running in delays due to the change in day light saving. These efforts resulted in reduction of tickets, reduced turn around time to the customers, build confidence within the stakeholders and also enabled my employer to establish better client engagement.

    • Colleen,
      I know all too well the importance of keeping a paper trail, especially in a field where turnover rates are incredibly high. The last organization I worked for was formed 40 years ago. It was only about 5 years ago that the staff who had been there originally or were with the company for 30 years began to retire. Since most of the staff had been older they kept paper records of everything, but they each had their own filing systems. Some staff printed every email or had client files in own drawn but grant applications for that client in another file. The problem is that when they left, no one knew where anything was or how business and clients were handled. The data warehouse only began to collect “clean” data 2 years ago! Sometimes you have to work only with what you have and start implementing processes and procedures that ensure information is stored and accessible once employees leave to avoid such problems.

  • 1.Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?

    A recently acquired pediatric practice acquisition was scheduled for EMR go-live 3 days out. They were switching from a less pediatric friendly EMR. Myself and the project lead were scheduled to go to the location and perform round trip testing on all the computers and printers swapped out by the hardware team. It wasn’t clear on how long the hardware team would need to prepare all workstations prior to us (the application) arriving for testing. It wasn’t clear what applications needed to be installed or tested.

    The entire process followed an unnecessary trial and error workflow. We (the application team) would log in to a computer to test the access to the EMR application to only find that it (the EMR) was not loaded on to the computer. to resolve this issue, the hardware team had to install it. Windows printing needed to be installed on all computers to test network printing. This step was over looked by the hardware team as well.

    At the time I was unaware of how much disconnect there was in communication between the hardware team and the application team when it is obvious that the successful testing the first go round was dependent upon a good checklist of necessary inputs and outputs. The hardware team wasn’t so forthcoming of what their direct responsibilities were to the application team so they (the app team) could cross validate any potential missing components needed from the hardware team to seamlessly complete testing requirements.

    • Brinn, based on the scenario that you have shared there is certainly a lot of communication gap that existed between the two teams. I believe there is a list of actions could have been taken in order to mitigate the problem that was created. First, the hardware team should have checked the old system to see what all softwares/applications were present in that system. Also, that list should have been shared with the application team to get a confirmation whether any additional applications were needed or not. After confirmation, the hardware team should have gone with the required installations. On the flip side, application team must have communicated before going to the location (regardless of the actions taken from hardware team) that whether the basic and obvious systems/applications that are required to do the testing were in place or not.

    • Brinn, I agree communication is many times the key to why project launches are unsuccessful. It seems techs and developers (and most people) can become so focused on their individual contribution that they lose sight of the fact that in order for the system to work, the teams have to work together. I’ve worked on so many projects where one team fails to communicate important details to the other team because they might not find them important. In the program, in particular, I’m learning there has to a balance between personal accountability and responsibility but also relying on others to get the job done correctly.

  • I had one very recent experience about unclear requirements, I had to develop a reporting framework with an unknown number of parameters for every report , they had to get configured through the DB and the program I had to write had to figure out how to extract the parameters and call the reports and the parameters validation as well. Hundreds of reports kept on coming and the code was looking like a huge mess somehow was working but had to do a lot of patching.

    • I curious as to how you or others determined the parameters to use and how the patches that were used got determined. Was there any QA points that took place or data gathering that helped produce the necessary outcome?

      • Every report has its own BRD about the parameters and new reports kept to get added on the fly, while the framework was done upfront. The patches were addressing the things which were added later and not having been counted in the initial design.

  • Watson is quite not ordinary software development system and doesn’t have much in common with the average development shop. All Software development systems have requirements defined or almost defined as needed, while in the Watson case they can flex and change at any moment. The IBM has invested a lot of R&D resources and money in the project combining NLP, Statistical machine learning, knowledge representation and reasoning, info retrieval. An expertise of several sciences was integrated in order to make the project to work, it’s hard to afford it from economical point of view as well..

    • I agree. Watson was not an out of the box computer. There were so many open ended correlations that had to be determined so the robot could make intelligible responses within seconds based on puns. This type of algorithm was out of the ordinary and difficult to culminate in the expected timeframe needed. I wonder if this technology could’ve been completed sooner if the meeting of the minds took place sooner.

  • I got very used to having unclear requirements in my previous role in event planning. For example, my boss would often hand me a file for a party he just booked and expected me to plan the event from soup to nuts without having a conversation with him or the client. In each file there was typically a proposal (although many times it was not finalized), some email correspondence and cryptic notes that my boss had written. Many times the information on the proposal was incorrect (especially the timing and guest count) and I couldn’t make sense of the cryptic notes. Additionally, after I had a conversation with my boss about the particular event I would learn that some of the information (timeline, flow of the event, linen choices, etc.) he gave me in the file was incorrect, and other items he would make up on the spot. I quickly learned that it was important for me to be at client meetings in order to take my own notes and talk to the client about details my boss would forget to ask. Soon my boss realized how many details he let fall through the cracks and rarely had a meeting without me.

    • Good example of the importance of communicating requirements. It can be very frustrating to be in that situation. Not only did your boss do a poor job of communicating requirements to you, but it sounds like he did a poor job of eliciting requirements from the stakeholders as well. When I was in charge of running golf outings (which is essentially event management) in my previous role, I found that the more “annoying” clients who had a lot of demands and asked a lot of questions would tend to have better events and would be more satisfied because they were clear regarding their requirements. On the other hand clients who didn’t ask enough questions still had requirements, but didn’t communicate them effectively and when it came to the day of the event their vision and our vision could be at odds. It is very important to ask questions and elicit requirements from both sides to establish clear goals.

    • Lauren,
      I understand the frustrations with being given a project or assignment without any explanation and incorrect information. It becomes evident that the person handing it off didn’t think it was important enough to have the right information, but then it becomes important once it becomes your responsibility. At least you were given the opportunity to attend the meetings to get the information yourself. I’ve been in similar situations where I have been handed a project with missing or incorrect information and I was not allowed to attend the meetings surrounding it. This was because my boss wanted me to do the work but take the credit. I think instances like this take a great deal of perseverance and creativity in finding solutions.

    • Lauren, this kind of sounds funny as it reminds me of my time at my last job. I had already worked for more than a year at my internship with my manager. When I joined this new organization, I was looking for something challenging. I was probably a week or two old in the company, and my manager handed me a legal agreement. The agreement was between our company and one of the internet giants. It was my first time looking at an agreement, and I remember how I struggled from one cabin to another, to figure the legalities of the business. Because I had minimum to no background on the deal, I had to read the 50 page agreement word by word. Although in this scenario, the output was very specific, as the negotiations began, the pendulum kept swinging to and fro. So in that case, the output became somewhat vague, because there was bidding involved, and we as the client were fighting for every cent that we were offering to spend. Finally because both organizations were so stuck on what they wanted to offer, the deal fell through. I took it as a learning curve, and decided never to deal with those agreements (especially when I had to read them). Till date, I haven’t figured if my manager was only trying to test me, because the deal was ridiculously in favor of our organization.

    • Lauren, a very simple and yet very relatable example. I think almost everyone can relate at least one of their incidence to this example. This is the case where we can point out the difference between the requirement gathering process and the requirement elicitation. I guess what your boss was trying to do was gathering the requirements and that too poorly or incompletely and what you were trying to do was to elicit the requirements. You were trying to reach to a conclusion by the hints or through inferences by the documents and notes provided by your boss, which is an efficient way to deliver products. But you realized and switched to getting yourself involved in the client meetings and started eliciting the requirements which were accurate information that made sense and helped you deliver efficiently and accurately.

    • I think communication and taking time to address the vital parts of projects are essential. I do also think that some things are better taught by the individual going through the process themselves, however the basics or any questions should be addressed. Without conversation it seems really difficult to get things done. At the end I do not think one should rely too much one anyone even if they are bosses like you explained Lauren and it is good you took responsibility of being at the meetings and following up with what you needed to. If one wants to get the job done correctly communication is vital, but if others are not wanting to take the time to do that then, one must do what they can with whatever resources they have.

  • Lauren, I certainly believe that is a nice approach to capture the requirements while getting involved in the meetings with the client. Just to add to that, circulating the requirements with all the parties involved would also help to get confirmation from client’s end and also ensure that everyone is on same page. In addition, this will enable to deal with clients who change the requirements quite often.

  • (1) A problem I was faced with was to report the correct number of HIV tests being administered to family planning patients. This may not seem like a problem, but figuring out how to do this would require not only months of investigation, but meetings between federal regulators, hospitals administrators, and internal managerial staff and IT support. I worked as a quality improvement coordinator for an organization that monitored the HIV testing of family planning patients as a art of a public health project. I had 10 health care centers on the project and some of them updated their data on a daily basis while others were having internal billing issues so they sent me aggregate data at the end of the month. Not only was our numbers not correct, we could not figure out why because there was no streamline system to follow or track. I started by meeting with each provider over the course of 3 months to try to trouble shoot their data issues to improve reporting. Once that was fixed, our numbers were still wrong. I then had to have meetings with the federal regulators who administered our grant funding to figure out if what they needed matched what we were collecting. I finally figured out that the data we were collecting was not in a format that fit their reporting requirements. Each provider administered their tests differently; some one test at a time, others running 2 types of HIV tests with one collection but then some billed twice while others billed once. We were also unable to know if a test was a first test or a confirmatory test. This was a problem that had so many working parts that once I left the organization, we were still working on it. The solution became clear, to mainstream how providers collected data and submitted it to us in a way that fit the reporting requirements. How to do that is unclear to this day and is still a work in progress.

    • This seems extremely complicated- but defining the problem seems to be half the issue. Approaching these complicated, multiple moving parts issues seems to depend on the definition of the issue. It seems as though your company saw an issue, but struggled to pinpoint the exact problem. Using the root-cause analysis we discussed last class may be helpful in this situation.

  • 2) There is an initial difference in the approach IBM took in solving the Jeopardy problem, but the general process does not seem to be so distinct from other processes we have seen. I think their first issue was how to approach the problem. Because this was an issue that had yet to be undertaken by programmers extensively in the past, IBM’s initial approach had no boundaries. In that respect, they had to be more flexible in defining the problem and coming up with initial solutions. However, as they progressed and refined the issues and solutions (defined their boundaries), they reverted back to a ‘normal’ approach: trial and error. They would create a trial program, test it, and refine the program. This is not an unusual approach. Their uniqueness came at the beginning of the process in trying to redefine a unique problem.

    • Margaret, when you say trial and error as the normal approach, can you give any example on that. What i understood from the case was that the normal approach that they referred to was the traditional SDLC approach that most of the IT developments follow. In this approach, you generally get the requirements which are defined and then you build the solution based on the requirements and after the tests it is deployed. But in case of Watson, the first step, that is the requirements itself was not defined. Hence, it cannot follow the normal approach. The requirements had to be developed real time while it was playing the game. And to fulfill this, there were combinations of many algorithms that were used. According to me, Watson couldn’t have been a success or even possible if it went through a normal traditional SDLC approach as it involved cognitive learning.

  • At my previous job working as a counselor for adults with disabilities, we often had issues with our clients growth and development. Despite having a defined diagnosis, as are the rest of us, each individual was unique and had individualized obstacles. Often, finding ways to aid them in growing and developing intellectually and socially was difficult because it was difficult to communicate with them. For example, one person I worked with was autistic, and communicated differently than others. To help him grow, I first had to dissect how he was communicating. After I learned the context in which he was operating, it was easier to develop ways to assist him. This was a long and tedious process that involved communication with his family and a lot of trial and error. After I was able to create certain ‘boundaries’ of comprehension and communication, I was able to develop solutions for him that worked within his capabilities.

    • That’s a really good approach that you followed. That’s true. It’s often difficult to have a specific or definite rules in place because the way people react is different and hence different approaches needs to be used. I believe it is quite challenging to be a Counselor as he/she has to really understand a person’s mentality in order to coach/guide him. Trial and error requires a lot of effort, lot of permutations and combinations are to be applied but sometimes it turns out to be the most feasible way to tackle problems.

  • 1. Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
    I took a job with a biotechnology startup company in Boulder Colorado. The founder was a professor and 1/2 of the 10 employees were his former students. The project was to synthesize RNA sequences which had been identified on SELEX technology to poses potential medical benefits. There were only commercial protocols to for some methods and the rest were either unknown or in some academic lab notebook. We had funding for 2 years and out first proof of concept in 6 months. The staff was from different areas, some were chemist, a few molecular biologist, an MD in pulmonary diseases, and a historian. The idea was to work in teams as rapidly as possible to prove the methods worked and try to move the project forward in two weeks. After two weeks they evaluated the direction of the project & if something more exciting came up you were moved to work on the hot area. We had to learn to communicate in using terms we all could understand and make it happen at the end of the day.
    The only know was that at the end of the day you could drink beer at your desk. It was a dynamic and challenging environment where priorities moved quickly and you worked on Saturdays. Needles to say, we found ways to get things done, move projects forward and secure more funding by the 3rd quarter.
    How did we figure things out? Trial and error and plenty of discussion on how to fix the problems. Also work with vendors who were eager to show us the products they sold and see how we could use those technologies to address and move the project forward. Most recently Merck bought a HepC that our group licensed out to another biotech.

  • This was the time when new projects were flowing in and my delivery unit was moving into agile methodology from waterfall method. Before we started implementing agile approach on the projects, we went through trainings where we were taught through examples. But as said, that one truly learns only when they have to apply their learnings in real life scenario. When the requirements came in from the clients, there were two people who were assigned from the off-shore team to work on them and eventually design screen wireframes based on that. Among the two people, i was the senior one, which means that i have to be the point of contact with the onshore team and the clients. Since this was the first hand on experience in agile and we were told that it will run in sprints and the we should expect changes in requirements, we were ready for that. When the first gathered requirements were given to us, we started working on them, although there were at certain points where we thought that there were gaps, but our architect convinced us that it will get cleared as we move further. But as we moved through scrums and calls and meeting, the requirements seemed more ambiguous and developing screens based on that were getting even difficult. The flow of the screens didn’t make any sense and if we thought of how to code them, it was becoming even complex. The backlogs started increasing. At that point our manager decided that we will stick to the waterfall method for the requirements process and implement sprints in the development phase. We decided to have versions of requirements, and only start developing when a version was signed off by the clients.

    • Hi Snigda, thanks for sharing. I found new initiatives are less efficient in getting the job done. The positive in your case is that you had a sound validated method to get the job done. I wonder if the client was operating using the Agile methodology or not. Also working remotely in different locations can be challenging when you are trying out a new approach. When I was in pharmaceuticals often time we had to work on collaborations in China. When we ran into problems trying to understand where we wrong was challenging. One time we were running different versions of the same software and it took us a while to figure out that was the problem. And yes, we were behind schedule because of this oversight. Best

  • 1.Give an example of a problem you’ve faced (from your own experience) whose solution had unclear requirements. How did you go about figuring out what the requirements really were?
    I do not have any work experience to answer this question, however in my life experience when I had to find an affordable healthcare insurance plan from Obama Care, I think it had very unclear requirements and solutions to finding a health insurance, at least for me. The website was very unclear and not helpful and when filling out the forms the wrong information would appear in reviewing the application many times. After many many trials there was no good end. After calling by phone and after many months of waiting a solution was made in finding a good insurance plan.

  • This is a very good example of a complex issue with an unclear solution. I have heard horror stories about uninsured people attempting to sign up for Obama care and how inconsistent the process was in shopping around from the correct insurance to fit your needs.

  • This is just for fun reading : IBM’s Watson should run for president

Leave a Reply