Home/
Part XV — Troubleshooting, Checklists, and Reference/48. Glossary and Quick Reference/48.4 Architecture patterns for LLM apps
48.4 Architecture patterns for LLM apps
Overview and links for this section of the guide.
On this page
Patterns
- Router: One small model decides which big model to call.
- Evaluator: One model generates, another model grades the work.
- Map-Reduce: Summarize 10 documents individually (Map), then summarize the summaries (Reduce).
- Reflexion: Model generates -> Model critiques self -> Model regenerates.