From hackathon to AI-901
June 20 is not the end. Each hackathon artifact maps to a second-semester AI-901 lab. This page shows the bridge so teams know what to keep and what to rebuild.
Hackathon team assignment becomes the model for second-semester Foundry project groups.
June hackathon
Each team uses one Foundry project (fdry-koicatiu-teamXX)
Second semester AI-901
Same project is reused for Second semester mini-labs and the team prototype sprint
The same endpoint, deployment, prompt, and .env pattern reappears in AI-901 labs.
June hackathon
Hackathon teams call gpt-4o-mini or the assigned model from Python using AzureOpenAI client
Second semester AI-901
AI-901 Lab 2 (Model Deployment) reuses the same code as the entry point
June teams iterate on system prompts; Second semester students treat prompts as code artifacts (versioned, tested).
June hackathon
System prompt lives in prompts/system.md, tested against demo questions
Second semester AI-901
Same prompts/system.md format used for AI-901 Lab 3 single-agent prototype
Hackathon RAG demos seed the Second semester deeper RAG work: indexing, evaluation, citations.
June hackathon
Optional RAG sample with 3-5 public docs (TIU FAQ, syllabus)
Second semester AI-901
AI-901 Lab 4 (RAG) formalizes upload, retrieval, citation, evaluation
Hackathon safety checks become lab checks for hallucination, privacy, cost, prompt injection.
June hackathon
Risk register section in team README (5 lines)
Second semester AI-901
AI-901 Lab 5 (Responsible AI) expands risk register into a full review
Working June demos can be refined into Second semester semester projects or portfolio artifacts.
June hackathon
Demo + slide + README + screenshot/video backup
Second semester AI-901
Project sprint week 4 converts a hackathon sample into a polished prototype
Demo + slide + README submitted. Risk register present.
Operations team reviews cost, completion, tutor feedback. Second semester quota recalibrated.
Returning teams keep project access. New students get fresh accounts.
Lab 1 — Foundry project setup. Hackathon participants act as peer guides.
Lab 3 — single-agent prototype. Teams pick a real project they will keep.
Lab 4 — RAG. Teams convert their hackathon sample into a grounded version.
Semester capstone. Teams ship the refined prototype.