·2 min read
Why parent-child chunking beats flat RAG on dense documents
Fixed-size chunking fragments technical and legal documents. Parent-child indexing preserves structure and improves recall without bloating the index.
RAG · Retrieval · Python
Writing
Notes
Technical notes on retrieval, agent infrastructure, and shipping production AI systems.
- ·3 min read
Async memory tiers for long-running LLM agents
Flat vector stores treat every memory equally. Tiered memory with promotion rules keeps hot context fast and archival recall bounded.
LLM · Agents · Python
- ·3 min read
Shipping LangGraph pipelines without a full MLOps stack
Multi-step agent workflows need orchestration, but most teams don't need Kubernetes. LangGraph state machines plus FastAPI webhooks get you to production faster.
LangGraph · FastAPI · Agents