Convergent Build Teams: A Mid-Semester Look Back

Joshua Pham
Texas Convergent
Published in
8 min readApr 4, 2018

At Convergent, we’ve been offering a semester-long program open to students across UT Austin. Each build team brings students through a product development curriculum and project-based learning experience, providing them with an opportunity to create software products across disciplines in much the same way as they would in industry. As Rainier described in his article, we knew by the end of our first semester, Fall 2017, that we had struck a chord through the build team experience. Students from CS/ECE to Business — to even International Relations — formed unlikely project partners to build products with market analyses, UI mockups, agile methodologies, and working prototypes. But build teams were both a semester-long commitment and a large-scale experiment, not just any weekend hackathon or a class project. There was plenty of room to improve the structure. So our incredible leadership team made changes for this Spring semester, towards enabling members to complement each others’ skillsets and maintain steady progress over the semester.

Last semester, we designed build team around problem statements, which we crafted before the start of the semester. Instead, this semester we selected technology areas: Business Process Automation, Health Tech, Natural Language Processing, and Blockchain. We left it up to the members to choose problems within those areas to solve.

Kickoff!

Students submitted build team interest forms indicating their level of interest in each of the technology areas, their level of comfort with areas like data science, development, and financial analysis, and friends they’d prefer working with. Based on the responses, Convergent leadership organized members into the four build teams in a way that distributed their skills and interests.

Distribution of interests among Build Team members

With palpable energy and excitement, each of the build teams began the program with a Kickoff. It was in these 3-hour sessions that we laid out the big picture and logistical groundwork for the semester, fostered camaraderie among like-minded students and diverse backgrounds, and asked dozens of “what-ifs” and “whys” as we brainstormed potential projects.

In these Kickoff sessions, we leveraged what we learned about brainstorming from the Fall, each build team with its own twist. Last semester, we had introduced design thinking as a build team workshop halfway through. This semester, we used it as a tool to lead the charge. During the Health Tech Kickoff, students contributed their own Health Tech challenges on post-it notes, which were then organized into clusters on a whiteboard, each of which was summarized into concise problem statements. Then, following a design thinking workshop devised by the Stanford d.school, the students paired up to flesh out problems as personal experiences and craft potential solutions through empathetic discussion. The Business Process Automation Kickoff, on the other hand, used a shorter exercise inspired by the d.school’s workshop to guide the brainstorming session. There, students mingled to talk about what they wanted to work on in Convergent, and used design thinking to formulate new ideas. In this way, they formed preliminary teams organically.

Problem area clusters for Health Tech (left); Business Process Automation Kickoff (right)

Creating momentum was another important goal of the Kickoff events. We wanted members to leave the sessions with a clear way forward, a sense of chemistry, and a healthy dose of hype. To keep things going, we onboarded all of our members onto Slack (a communication tool with powerful filtering features and integrations) and settled on a weekly meeting time.

“By the end of the kick-off, our build teams already had a strong sense of direction and were HYPED about their projects” — Luca Tomescu, Blockchain

Reflections at the halfway point

By this time in the semester, build teams projects are well underway. Each build team consists of 15–20 students, working on 3–4 different projects. We’re all rewarded by what we’ve seen — from students who feel personally connected to their projects, to challenges and realizations faced because of collaboration across disciplines.

Students at UT Austin are familiar with SURE Walk, a campus initiative to accompany students on their way home if they feel unsafe, ever since it began due to a campus homicide in 2016. In Business Process Automation, one team called “Sure Run” is tackling inefficiencies in the matching process that has caused students to wait too long or not be picked up at all.

Texas Convergent is helping The University of Texas become a safer place. By connecting different majors, we have seen [Sure Run] grow stronger as it is shaped with user-centric design and has a market analysis and valuation. — Martha Czernuszenko, Business Process Automation

SureRun team presenting application flow for a weekly update

Another project, arising from a Health Tech build team, aims to improve the way youth are educated on the effects of alcohol, by developing an augmented reality mobile application for Google Cardboard. The team seeks to emulate alcohol impairment simulation goggles from middle schools, supplanting them with a more cost-effective solution and informative tooltips built into the immersive experience. To top it off, the team is building upon background research from one student’s class.

The mission of Convergent build teams is closely tied to what our leadership wished had existed before getting our first internships: the experience of building something together with others that contribute different perspectives and skillsets. With build teams, we want to enable students to enter future internships with an understanding of the value of their education, as well as with motivation grounded upon the desire to create.

One team lead describes how interdisciplinary skillsets have rubbed off on each other, in order to clarify goals of the project:

“Team members traditionally focused on providing the best engineered solution begin to consider user-centric design, and team members focused on user-centric design began to reflect on how user features could be built into the proposed architecture. For example, one of the teams, self-named “0–3,” has gone above and beyond the basis for the UText case. Instead of just integrating keyword search, as Student Government has requested, their interdisciplinary approach to the root problem has helped them identify that this tool is only as useful as the quick search is efficacious. They identified that most interactions with UText would occur when a student quickly needs information; in that situation, it is highly unlikely that they will remember what keyword will bring up a service’s information. Knowing this, they are in the process of broadening the range of keywords by integrating a lexicon.” — Arnav Jain, Natural Language Processing

On maintaining momentum

14 weeks is a long time to devote to a project. This time, coupled with other commitments, studying, and projects, between leads and members alike, means that strategies to keep the momentum are extremely significant to have. But members are not alone; they benefit from working on their own projects as part of a large organization. Leads from across build teams share various methods used to keep a pace of progress throughout the semester. Several themes emerge: planning ahead, and setting up week-by-week plans. Leads borrow many of these techniques from organizational skills learned in past internships, such as agile methodologies like sprint cycles and productivity tools like Trello.

“Whether that’s using a Trello board, Google calendar, or a any organizational tool, by putting it “on paper”, you can make sure you have the time to commit to your project.” — Martha Czernuszenko, Business Process Automation

“A few weeks into the semester, when our build teams had solidified their ideas and were just starting to think about developing their concepts, we had them lay out a week-by-week plan for the rest of the semester.” — Luca Tomescu, Blockchain

“At the end of a sprint, you have expectations of what the product should look like. Because you’ve planned ahead, you also have an idea of what the next sprint needs to accomplish based on your expectations. If the sprint fails, if the specifications of the final product change, you must be able to adapt the sprint planning based on the sprint result. Failure to plan sprints ahead of time creates unproductive gaps which leads to a loss in interest.” — Shay Sayed, Natural Language Processing

Planning ahead

With experience, leads also start to recognize ways they can best support their build teams. From last semester, we learned that one effective way to do this was to assign a lead to each team that works on a different project. With 4 leads taking care of each of the technology areas, we divide ourselves to keep each of the teams accountable. Another way to look at this is making ourselves available to assist, de-stress, and mentor. As Martha puts it,

“Don’t be afraid to ask for help! The officers are here for you — whether it’s looking through code, taking your idea to the next step, or just taking a mental break.” — Martha Czernuszenko, Business Process Automation

With a few weeks left to go, build teams are in their final stretch. We’re all proud so far, whether it’s our ideas and execution, learning to accomplish a plethora of firsts, or managing to get the most out of Convergent build teams alongside our other responsibilities as students. We’ve heard one member proclaim that “Convergent is bliss,” another confess that they should be studying but instead are attending their weekly build team meeting, and yet others rave about the heaps of progress they made over Spring Break. And we’re all looking forward to Demo Day, on April 25, when all build teams get to showcase their projects. So stay tuned!

Joshua Pham, Architect for Health Tech build teams. Thanks to all the officers who contributed to the blog post, and to everyone who has made build teams and this blog post possible!

--

--

Joshua Pham
Texas Convergent

NLP & Recommendation @Spotify. UT Austin CS, 🏠 bay area/cambridge, MA