Grimoire ML

An AI-assisted Magic: The Gathering card discovery and deckbuilding app. A mono-repo of four services: a FastAPI backend, a query microservice that turns natural language into an intermediate representation and then into Scryfall queries, a Redis-cached card image proxy, and a Next.js frontend, with Prometheus and Grafana wired in from the start.

• A query service that goes natural language, then intermediate representation, then Scryfall query, with Redis caching between stages.
• A dedicated card image proxy with its own cache and metrics, keeping image traffic off the main backend.
• A Next.js 15 frontend in TypeScript with Tailwind and Mantine.
• Prometheus metrics and Grafana dashboards alongside a Docker Compose stack wiring Postgres, Redis and every service together for local development.

• An intermediate representation between the model and the API

  Asking a language model to emit Scryfall syntax directly is brittle and hard to debug. Going via a structured IR makes the model's output inspectable and lets the query layer be tested without the model at all.

• Cache aggressively at every boundary

  Card data barely changes and model calls cost money and latency. Redis between each stage made the app feel instant on repeat queries.

• Observability from day one on a side project

  Wiring Prometheus and Grafana in early cost an afternoon and paid for itself the first time a service started misbehaving under load.

Translating natural language into a structured intermediate representation before hitting the search API, rather than asking a model for the answer directly, is what made the results trustworthy.