Codematic AI/ML Engineer Assignment

Options 1 & 2 — lesson pipeline, intent classifier, offline LLM design

Links

Stack: Gemma 3 1B QAT · Ollama · FastAPI · Cloud Run

API status

Checking…

Option 1 — Intent classification

Typical latency on Cloud Run: ~8–11 s (warm instance).

Option 1 — Lesson plan generation

Typical latency on Cloud Run: ~3 min for CUR-NG-001 (sequential LLM calls). Leave this tab open.

Option 2 docs (in repo)

See the docs/ folder in the GitHub repository for full Option 2 deliverables and architecture charts.