LangChain Blog (Page 8)

Few-shot prompting to improve tool-calling performance

We ran a few experiments, which show how few-shot prompting can significantly enhance model accuracy - especially for complex tasks. Read on for how we did it (and the results).

By LangChain 8 min read

How Athena Intelligence optimized research reports with LangSmith, LangChain, and LangGraph

See how an AI-powered employee for enterprise analytics used the LangSmith playground and debugging features to quickly identify LLM issues and to generate complex research reports.

Case Studies 4 min read

Planning for Agents

The fourth installment in our "In the Loop Series," in which we talk about what planning means for an agent and how to improve it.

In the Loop 6 min read

Improving core tool interfaces and docs in LangChain

See our latest improvements to our core tool interfaces that make it turn any code into a tool, handle diverse inputs, enrich tool outputs, and handle tool errors effectively.

By LangChain 4 min read

How We Deployed our Multi-Agent Flow to LangGraph Cloud

Read this guest blog post on how to create a LangGraph multi-agent flow via React & LangGraph Cloud.

5 min read

Why you should outsource your agentic infrastructure, but own your cognitive architecture

In this third installment in our "In The Loop" series, learn why you should tailor your cognitive architecture to be application-specific, alongside running better infrastructure for your agentic apps.

Harrison Chase 3 min read

[Week of 7/8] LangChain Release Notes

See new use cases for building & deploying with LangGraph Cloud, revamped LangGraph docs, and more on self-improving evaluators in LangSmith. Plus, hear the latest agents trends — and learn more about our upcoming agents-themed hackathon.

4 min read

LangSmith for the full product lifecycle: How Wordsmith quickly builds, debugs, and evaluates LLM performance in production

Learn how WordSmith, an AI assistant for legal teams, uses LangSmith across its entire product lifecycle — from prototyping, to evaluation, to debugging, to experimentation.

Case Studies 5 min read