What I Have Been Working On With Hult Prize

I started my Impact Fellowship with Hult Prize in February 2026. At first, most of the work was what you would expect from a startup evaluation process: watching pitch videos, reading pitch decks, going through application details, and trying to understand whether a team had a real problem, a believable solution, and enough traction to move forward.

That work became more concrete once the volume increased. I have now evaluated 80+ startup applications, which means I have seen a wide range of teams, markets, decks, and levels of clarity. Some applications explain the problem sharply but do not show enough traction. Some have traction but make the business model harder to understand than it needs to be. Some are not polished, but the founder has clearly spent time close to the problem. Going through that many applications made evaluation feel less like judging a single pitch and more like comparing evidence under time pressure.

The second stage added interviews. I have conducted 15+ national-stage online interviews with startup teams, and those conversations changed how I understood the applications. A deck can make a company look cleaner than it is, but it can also hide the strength of a founder who explains the problem much better in conversation. In the interviews, I had to listen for the same signals as the written review: market opportunity, traction, founder readiness, and whether the team understood what they were building beyond the pitch language.

I have also been helping on the technical and operations side. One of the first problems was simple but repetitive: sending similar emails to many startups, with small changes in each message. I wrote a Python/Selenium automation script for that workflow so the outreach could be handled more consistently without manually repeating the same steps each time.

The larger technical work has been around data and evaluation tools. A lot of the useful information sat inside Excel application data: traction, evaluator reviews, comments, scores, regions, and other fields that mattered to stakeholders. I built dashboards that made those data points easier to filter and compare, so the team could look across startups instead of reading everything one file at a time.

I also built an internal evaluation platform for the later stages of the process. The goal was to present 200+ startups in a way that made evaluation easier for reviewers as they moved from national selection toward the incubator stage. That work was not just about building a page or displaying records. It was about understanding what evaluators needed to see, what information should be easy to compare, and how to reduce friction in a process with a lot of moving parts.

What I have liked most is that the fellowship has not stayed in one lane. I am learning how startups are evaluated, but I am also seeing the operational work behind that evaluation: the dashboards, the communication, the data cleanup, and the tools that make the process usable for everyone involved. That combination has made the work more valuable for me because it connects judgment with implementation.