LLM Wikis and how to give your agents memory @LangChain
LLM Wikis and how to give your agents memory  @LangChain
Uploaded July 2026 | Updated November 2026, 1 day ago
LLM Wikis are emerging as a common pattern for giving context to agents. Examples include DeepWiki (Cognition), AutoWiki (Factory), and “LLM Wikis” (Karpathy).

LangChain recently launched OpenWiki, an open source project for creating and maintaining your own LLM Wiki. Dosu has been working on “knowledge infrastructure for agents” for the past three years.

Join Harrison Chase (CEO and Co-founder at LangChain) as he moderates a discussion between Brace Sproul (Head of Applied AI at LangChain) and Devin Stein (Founder and CEO at Dosu) as they discuss how to best build LLM wikis.

Topics will include:

What is an LLM Wiki?
Are LLM wikis even the right abstraction?
Are LLM wikis for agents or humans?
What is the hardest part of building wikis?
LLM Wikis and how to give your agents memoryHow Lyft Increased Its Agent Resolution Rate by 16% with LangSmith and LangGraphAutonomous Agent Improvement with LangSmith Engine | New LangChain Academy CourseHow to use dcode + Nemotron 3 UltraHow Airbnb builds Trust for Guests and Hosts using AI Agents with LangGraphTrace Every Cursor Agent Turn in LangSmithLangSmith: The platform to run, improve, and govern agentsBuilding a Production Agent Eval Pipeline: Harbor + LangSmith + OpenAI SDKOpenWiki 0.2 is adopting the OKF specCreate an agent that can browse the web with Managed Deep Agents and Browserbases StagehandBuilding Semiconductor Agents at Skyworks with LangGraph + LangSmithWhy Distillation Might Be Impossible to Stop
LangChain |

LLM Wikis and how to give your agents memory

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER