Harrison Chase
Also known as LangChain
Harrison Chase co-founded LangChain after publishing the original language model application package in late 2022. He and Ankush Gola started the company in early 2023. LangSmith followed in public beta in July 2023, with tools for tracing, testing, evaluating, and monitoring model-powered applications. LangGraph later addressed requests for more control over agent flows and state. Chase’s anniversary essays explain how user feedback shaped those changes, including problems with hidden prompts, dependencies, and reliability. His public relevance comes from original code and sustained explanations of how developers can operate AI applications. His essays connect LangSmith’s evaluation tools and LangGraph’s state management with problems reported by application developers.
- Now at
- LangChain
Timeline
-
Late 2022
An open source side project
Chase pushed the first LangChain code from his personal GitHub account in late 2022. The package connected components used in language model applications. It helped developers start experiments as they explored what the new models could do.
-
February 2023
Starting a company with a collaborator
Chase and Ankush Gola established LangChain as a company in February 2023. They set out to help developers build reliable agents and applications. The company grew from the open source project, with the team responsible for the later products.
-
July 2023
Making application behaviour visible
LangSmith entered public beta in July 2023. LangSmith provided debugging, testing, evaluation and monitoring across an LLM application’s steps. Developers could inspect model inputs and outputs rather than relying only on a final answer, and use those records to investigate failures.
-
2024
Giving agent builders more control
By LangChain’s second birthday, Chase described LangGraph as a response to developers who needed more control over agents. It supplied a lower-level way to manage application flows and state. Chase also cited reliability problems with preconfigured approaches and a need for clearer control of context.
-
October 2025
Reviewing three years of changes
Chase discussed package complexity, hidden prompts, dependencies and user requests for more control. He explained how LangSmith became separate from LangChain and how the team moved into LangGraph. These discussions shaped the team’s successive tools.
-
Current public role
Explaining agent engineering
LangChain offers an open platform around agent engineering. Chase continues to publish explanations of the field and the tools his team builds. His role connects an original software project with a wider public discussion of evaluation, context, and reliable AI applications.
Sources and credits
Photo: Harrison Chase / GitHub profile. License: Not stated by source. Source: github.com. Cropped to a square and resized. Ask us to remove or replace this photo.