ScholarComp
A results platform for K-12 academic competitions, built on more than 13 million records from thousands of inconsistent sources.
- Sector
- K-12 education
- Work
- Data platform, product, web
- Records
- 13M+ results
- Team
- Two people, with agents
The problem
Academic competition results scatter. A regional science olympiad posts a PDF. A debate circuit runs its own tournament software. A math league emails a spreadsheet. A district publishes a press release. Nothing connects, nothing persists, and a student who competed for six years has no record of it anywhere.
Athletics solved this decades ago. Academic competitions never did.
What we built
- A results database at scale
- More than 13 million competition results, ingested from sources that agree on nothing: different formats, different naming, different levels of completeness.
- Entity resolution
- Schools appear as “Centennial HS,” “Centennial High School,” and “Centennial Senior High,” in districts that reorganized. The matching system decides which records describe the same school, with confidence scoring and a review path for the hard cases. Everything downstream depends on getting this right.
- A rating system
- A hierarchical rating model that places results from competitions with different fields, formats, and difficulty onto a comparable scale.
- A school database
- School records sourced from federal data, joined to competition results.
- Event-driven services
- Independent services with isolated data, communicating through an event stream, so ingestion, matching, rating, and publishing proceed without blocking each other.
- A data warehouse
- A layered pipeline feeding reporting and analysis.
- Mobile-first front end
- Students and parents check results on phones. The site starts there.
How we built it
Two people direct the platform. Agents write the code against written specifications, with automated gates and independent review passes enforcing an acceptance standard committed to the repository.
A traditional team building this would have needed a data engineering group, a platform group, a front-end group, and the management layer that coordinates them.
Where it stands
ScholarComp runs in production with the results database, matching pipeline, rating system, and public front end live. Organizer tools are in development.
Visit scholarcomp.comTell us what you need.
Send a paragraph about your organization and the problem in front of you. We'll tell you whether we're the right shop, what we'd build, and what it costs. If we're not the right shop, we'll say so and point you somewhere better.