Sean Devlin, contacted Net Solutions in 1997 to build a flash website. After that, he joined Monster.com and worked on some cutting edge features for them. After almost a decade Net Solutions heard from Sean again. Sean had partnered Brad Downs to work on a revolutionary online product, which has subsequently rewritten how universities across United States manage their sport athletes, recruits, alumni and events. The product is Frontrush. The success story continues with many records broken and accolades earned. Net Solutions, a part of this phenomenon right from the inception, is proud to be the development outsourcing partner for the premier University Sports Management System.

The project started as fixed price project in early 2006. During the pre-sales stage itself, we did some rapid screen mock ups and quickly went to development. Manoj Kalra, who had joined Net Solutions a few months back was the lead developer for the project. The project was developed much ahead of schedule and we were discussing Phase II of the project in May 2006 itself. While the first phase of the project was mostly about creating basic functionality for coaches and assistant coaches to manage recruits, rosters, notes and duties - the phase II added the calendar for events/ duties/ notes as well as stronger administrative control. Like Phase I, Phase II was also completed ahead of schedule and in late 2006, we were already to Phase III. Phase III focused on personalization of each coaches area with custom form fields, stronger search features and interactive AJAX features. As Phase III was being developed, the product was under extensive testing and demo at target universities and through some coaches. It was also the time when the first sales were made. Up till Phase III, we had worked on a fixed price approach. We quoted for each Phase, delivered a project plan and delivered much earlier. This was in large part due to an effective product owner role played by Sean. Sean made himself available at India times, almost every time we needed him to be available. He gave feedback quickly, helped us work in short weekly cycles, maintained a very clear prioritized requirements roster based on approved features for each phase and keep our focus on particular week’s deliverables. All the three releases [I, II and III] were delivered without any priority 1 or 2 bugs and much ahead of schedule. The application was beginning to take shape.

It was roughly around a year since Net Solutions and Frontrush had been working together. Also, the product was now in the market. The changes, customizations and support requests for each university were different than others. In addition, strategic liaisons with third party products like email template makers were being explored. At this stage, Net Solutions introduced the idea of abandoning Phase wise approach and working with a requirement roster for only few weeks. Sean agreed to try out the idea. Sean agreed to the idea and discussed the same with Brad as well, who seemed excited about the idea. Net Solutions history of good deliveries and retaining the same team only helped make the decision easier. While coming to Phase III, we had already demolished the hierarchical structure of project management from the project altogether and Manoj was the primary contact for the project for Sean. Going forward, Manoj was primary contact for Sean with a Business Analyst helping Sean define requirements/ acceptance test criteria as needed. The first test of this engagement was a delivery 6 sprints later. A lot of functionality - search, performance, usability and also a critical email template integration for emails being sent from Frontrush by various coaches was being deployed. And we were able to do this on time. After this, the team really took off. Sean and Manoj were on MSN almost daily discussing priorities, checking features and making each sprint a success. The whole of 2007 saw Sean and Brad travel across United States and add one prestigious client after another. The development team in India supported them through quick turn around of demo requests, feature integrations with existing university websites as well as customizing any parts of Frontrush as needed. In late 2007, the team grew even more as we worked on releasing yet another Frontrush product - Camper : a solution on the lines of Frontrush, but targeted at recruit screening camps. The team once again pulled off the product and integration in less than 03 months. In the meanwhile as Frontrush continued to grow, we also worked on support, performance tuning and upgrades.

Frontrush has been a delightful journey for all involved. Net Solutions holds Frontrush as one of its premier projects and is working with FR team in US to now increase the market lead even further.

Popularity: 10%

Posted on March 11th, 2008 by admin
Filed under: Case Studies, Collaboration | No Comments »

MingleNow.com was one of the projects that started the Agile wave at Net Solutions. Ishita Chaudhuri, the project manager, was involved in the project from conceptualization in October, 2005 to its final day on January 07, 2008. In this email interview, she shares her experience in the world’s leading nightlife social networking website.

agilecollab: What has been the greatest learning personally and for the team, by working on MingleNow.com?

Ishita: MingleNow.com was a technically challenging project and we learnt many new things. The client for this project, Blue Lithium was one of the largest advertising networks in the world and as you would know they were recently acquired by Yahoo. We expected it to be fairly challenging but the combination of technical and business dynamics, made this a unique and complex yet rewarding project for all of us. Specifically, we learnt how to handle multiple server environment ie. how the code functions on a cluster of servers and optimization of code using memcache and pearcache. We also had a do a a little bit of unlearinig i.e. optimizing database by de-normalization. The team also got a chance to work on advanced perl scripting using image resing functions with Imagemagick. MingleNow.com was among the first projects to have a central code repository and use XP principle of shared code. We used SVN and created an automated environment for coding and deployment using SVN. We also did some automated testing using Selenium and load testing using JMeter. From personal view point, I think I learnt a lot about how to understand and respond to the customer’s expectation and how to maintain cool in a critical situation. It took some time to reach a stage where I could look at a critical issue and not jump out of my seat, instead thinking of ways to sort it out, having a plan b etc. And most importantly the team learnt how to organize itself to be responsive and accept change as a way of life for a new product development.

agilecollab: At what stage did you decide to abandon the idea to control the requirements and rather organize yourself and the team to respond to changing priorities?

Ishita: Around the time that we released the beta, we realized that there were lot of changes just after a deliverable. For instance, we changed the design of the complete website 04 times in 02 years. These changes were necessary for the site but were eating into the approved time for the project. This made us look for a solution that would be beneficial for both the client as well as for us. Hence, we went from a fixed price approach to an hourly approach borrowing some principles from Agile - like daily team meeting, collaborative code and a roster of prioritized requirements. Other things followed.

agilecollab: What sort of “testing” was implemented on MingleNow.com?

Ishita: All the team members tested their code before releasing it to the repository. There was no automated unit testing done however. At one point in time, the site got so advanced that it was not possible for the testers to test everything and hence, we added a mechanism for rigorous automated testing as well as dedicated team member for regression testing. Upon the request of the team [and conceded by the customer] we had three servers ie. development server (for coding), QA server (for checking final deliverables after merging) and live servers. We also had a big beta testing phase and lots of bug bashes internally which gave us insight into usability issues.

agilecollab: How did you work with “UI design team” and what are things you will keep in mind next time around?

Ishita: The UI design team was involved in all the discussions and hence were aware of all the functional requirements of a module and they also used the website as regular social networking users. They were encouraged to use all the sections of the site. This gave them an idea of the usability issues and they kept those in mind while designing for newer modules. Design concepts went through 4/5/6 and more iterations normally. These iterations were based not only on client requirements but also the development team’s feedback. After a while, the self organized process worked like a dream.

agilecollab: What was the biggest “team organization” challenge in a project that spanned over 2 years?

Ishita: The design team and development team worked at different hours so initially it was difficult to co-ordinate the discussion timings but then two designers volunteered to work as per the development teams hours and at each stage we had at least 02 designers working with us. This helped us a lot. Also, initially some of the development team members were used to working on single projects or maintenance projects only and hence it took them a little time to get used to the multiple developer scene but since modules were developed in such a way that at a time not more than two developers were working on a module, it was manageable. Another challenge was to work with the freshers in latter part of the project. However, the team members showed remarkable affinity to train, pair program and mentor the new resources.

agilecollab: Finally, how do you feel about MingleNow.com coming to an end and what is in store for the new year?

Ishita: Personally, I feel a little sad that MingleNow.com has come to an end because it was the first project that I worked on as a project manager and I would believe that I gave it all to make it work. It took a while for us to come to terms with the announcement. I also felt difficult to break the news to the team. When I did tell them, some of them also felt the same. It took a while to get out of the daily routine of chats with Krishna and his team for the first half of the day and then the mad rush of making the changes, reviewing, testing and getting it ready. The way users on the site took to the new features and their constant feedback kept us going and the constant growing number of users on the site pushed us more towards giving it our best.

Popularity: 16%

Posted on January 9th, 2008 by admin
Filed under: Case Studies, Interviews, Skills | No Comments »