Public case study for Frontier Build, an international startup-style AI competition for student builders.
Frontier Build brought student teams together to build functional AI products, test their ideas, and present five-minute investor-style pitches. Running the competition required infrastructure for registration, participant management, submissions, judging, communications, and technical support.
- 100+ teams
- 300+ students
- Participants across 6+ countries
- Judges, mentors, webinars, technical support, and external partners coordinated across the competition
- Winning participants were connected to internship opportunities
As Co-Founder, Technical & Operations Director, I worked across the competition's technical infrastructure and execution. My contributions included:
- building the public website and participant-facing competition infrastructure;
- building registration and submission workflows;
- maintaining participant data and administrative tools;
- building and supporting judging infrastructure;
- coordinating judges, mentors, webinars, technical support, and external partners;
- helping shape competition criteria around both technical execution and evidence that teams were solving a real problem.
[ Registration ]
|
v
[ Participant / Team Management ]
|
v
[ Build + Mentor / Webinar Support ]
|
v
[ Functional Product Submission ]
|
v
[ Judging Infrastructure ]
|
v
[ Five-Minute Final Pitch + Results ]
Frontier Build was designed around a simple constraint: teams needed to build something functional and be able to explain why somebody would need it. Judging therefore considered both technical execution and customer/problem understanding rather than treating either one as sufficient by itself.
That balance came partly from lessons I had learned while building Finora: technically working features do not automatically create a product people want to use.
This repository describes my own technical and operational contributions. Frontier Build was a team effort, and other directors, judges, mentors, partners, and volunteers contributed to outreach, judging, mentorship, participant support, and competition execution.
The original competition involved internal participant records and administrative systems that should not be exposed publicly. This repository therefore documents the architecture, role scope, and verified high-level scale without publishing private participant data.