Building Shoots: a hybrid event-driven agent system for photography
Description and use
Shoots is an automated photo analysis system that uses a hybrid agent and Pub/Sub architecture to ensure reproducible and transparent results.
Problem
Photographers need a system that can automatically review and provide evidence-backed feedback on their photos without relying on opaque model decisions.
Outcome
Shoots delivers a reproducible, auditable photo review pipeline that limits model authority, ensures safe retries, and provides evidence-backed feedback to photographers.
Workflow
- Upload photo → create durable Run
- Ingest stage processes media
- Analyst panel reads Shot with multiple readers
- Cartographer and Judge analyze results
- Scribe writes Shoot Record
- Optional Experiment offered
Evidence
The model panel reads a single Shot. The system around it does the work: it creates a durable Run, moves tiny events through independently retryable stages, re-reads state at every boundary, records every outcome, and only settles a Shoot after every member Run is accounted for.
I use Gemini 3.7 Flash for bounded visual interpretation and writing. Deterministic code owns evidence rules, state changes, and anything that sounds like a fact.
- Record date: 2026-09-07T04:52:25.389Z