Brace Sproul is Head of Applied AI at LangChain. He presents technical material about agents, including a walkthrough of OpenWiki, the organization’s open-source documentation system.[1][4]
Start with OpenWiki →
Links
X profile 
GitHub profile LangChain OpenWiki walkthrough Website Head of Applied AI · LangChain
Career
Sproul’s portfolio dates his LangChain Applied AI Engineer role to October 2023 and the team-lead role to March 2025. Before that he was a tech lead and full-stack engineer at Create, building a code-generation agent, and did contract development through Linear IO Services. He also founded the footwear and apparel distributor 510 Kickz. His open-source work includes maintaining LangChain.js and LangGraph.js.[5]
Contributions
Keep documentation connected to source changes
OpenWiki generates a connected wiki from a codebase and maintains source evidence for its claims. Its update workflow can revisit changed code and propose documentation changes, while the resulting pages can serve both people and agents. Sproul’s LangChain-published walkthrough explains the project’s ideas, operation and lessons from building it.[3][4]
Persist an agent’s progress between steps
LangGraph.js represents an agent as connected operations over shared state. Its persistence and durable execution support resuming long-running work; interrupts allow a human to inspect or change state before execution continues. Those mechanisms help an application handle pauses, failures and approval points without treating the whole interaction as one uninterrupted model call.[6]
Tools, writing & talks
Tools
Start here
OpenWikigithub.comAn open-source system for generating and maintaining source-linked codebase documentation.
Read how the wiki records sources and handles updates, then try it on a small repository you understand. Terminal and Git familiarity help; review generated claims against their linked code before relying on them.
- LangGraph.jsgithub.com
A TypeScript framework for stateful agents with durable execution, memory and human interaction.
Writing, research & talks
A LangChain-published walkthrough of the project’s design and operation.
Presenter[4]Original source