In the news
Retrieval
LangChain · Published · 3 min read
In 30 seconds
- What happened
- LangChain introduced a Retriever abstraction to support any retrieval method, not just vector databases, enabling easier integration of external retrievers.
- Why it matters
- Matters for engineers building question-answering systems or chatbots over custom data who want flexibility beyond semantic search or plan to use external retrievers.
- Watch out
- VectorDBQA chains still exist but are no longer actively developed; engineers should migrate to RetrievalQA chains, though the change is backwards compatible.
Listen to this summary
- retrieval
- retriever
- eval
- gpt
The patterns behind this
- Query Transformation Retrieval
- Hierarchical Index Retrieval (RAPTOR)
- Eval-Driven Development (Agent CI)
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.