Case studyHackathonSummer 2026
AskScotty: Campus Q&A Agent
A live campus Q&A agent built at a hackathon, orchestrating 24 AI tools through the Claude API.
AskScotty · App screensHackathon
TL;DR
- Role
- Owned product + engineering decisions
- Team
- 2 builders + 1 designer
- Timeline
- Summer 2026
- Tools & stack
- React Native · Expo · Django · PostgreSQL · Claude API
- Outcome
- Live; tested with 5–10 users
01 · The problem
[PROBLEM HEADLINE]
[THE PROBLEM: who has it, what they do today, and why it matters.]
[A USER QUOTE OR KEY INSIGHT FROM TESTING]
02 · What I did
Owning product and engineering decisions
Phase AResearch
[HOW YOU UNDERSTOOD THE PROBLEM AND USERS.]
Phase BProduct decisions
I owned the product and engineering decisions for the team. [THE CORE FEATURE CALLS YOU MADE AND WHY.]
Phase CDesign
Our designer [NAME] [WHAT THEY DESIGNED]; I [WHAT YOU BUILT FROM IT].
Phase DBuild
Built with React Native and Expo, a Django and PostgreSQL backend, and the Claude API. The agent orchestrates 24 AI tools. [ARCHITECTURE DETAIL]
[SCREENS / ARCHITECTURE DIAGRAM]
03 · Key tradeoff
[TRADEOFF HEADLINE]
[THE HARDEST TRADEOFF YOU MADE, THE OPTIONS, AND WHAT YOU CHOSE.]
04 · Outcome
Live and tested with real users
5–10
Users in testing
AskScotty is live. [WHAT TESTING SHOWED: [METRIC]]
05 · Reflection
What I would do differently
[REFLECTION]