Newsletter
Innovator profile
HC

Harrison Chase

Co-founder and CEO, LangChain

AI agentsLLM frameworksOpen sourceContext engineering

Machine learning engineering before LangChain

Harrison Chase worked as a machine learning engineer at Kensho Technologies from July 2017 to October 2019[1]. He then worked at Robust Intelligence from October 2019 to January 2023[1]. Fortune reported that he was an engineer at Robust Intelligence when he started LangChain[2].

Starting LangChain

Chase pushed the first code for langchain as an open-source Python package from his personal GitHub account in autumn 2022, as a side project[3]. He has written that he was inspired by meetups where developers were experimenting with language models, and that he packaged the common patterns he saw[3]. ChatGPT launched about a month later, and interest in the project grew quickly[3].

In the early months he added integrations with language models and vector databases, plus templates for tasks such as retrieval-augmented generation, SQL question answering and extraction[3]. He has written that helping users choose a model and later change that choice was important from the start, and that model neutrality remains one of the main benefits of the company's products[3].

Chase and Ankush Gola started LangChain as a company in February 2023[3][4]. It raised a $10 million seed round led by Benchmark in April 2023 and a $25 million Series A led by Sequoia in 2024[2].

LangSmith, LangGraph and LangChain 1.0

LangSmith, the company's platform for observability and evaluation of LLM systems, went into beta in summer 2023[3]. Chase wrote that it was built to be independent of both the model and the framework a developer uses[3].

In summer 2023 the team began receiving negative feedback about the original langchain package, including hidden prompts, breaking changes, dependency conflicts and outdated documentation[3]. The team began developing LangGraph that summer and launched it in early 2024 as a lower-level framework and runtime for agents[1][4]. LangChain 1.0, released in October 2025, focused on the core tool-calling loop and used LangGraph as its runtime[1].

On 20 October 2025 LangChain announced a $125 million Series B at a $1.25 billion valuation, led by IVP with Sequoia, Benchmark, CapitalG, Sapphire Ventures and others taking part[2][3]. LangChain's about page says the company works with 35% of the Fortune 500, has passed 1 billion open-source downloads and ingests more than 1 billion events a day on LangSmith[4]. The company is headquartered in San Francisco with offices in New York, Boston and Amsterdam[4].

Writing on agents

Chase writes about agent design on the LangChain blog[5][6][7]. In a June 2025 post he defined context engineering as building dynamic systems that give a model the right information and tools in the right format to plausibly complete a task[5]. He argued in the same post that most agent failures come from the right context, instructions and tools not reaching the model[5].

In July 2025 he described Deep Agents, an approach to longer, more complex tasks that combines a planning tool, sub-agents, access to a file system and a detailed prompt[6]. He pointed to Deep Research, Manus and Claude Code as applications that use this combination[6]. LangChain later released Deep Agents as an agent harness for autonomous, long-horizon tasks[4].

In an April 2026 post, Your harness, your memory, he argued that agent harnesses are closely tied to agent memory[7]. He wrote that using a closed harness behind a proprietary API means giving control of an agent's memory to a third party[7]. His AI Engineer speaker page lists him as CEO and co-founder of LangChain for 2026[1].

Known for

Sources

  1. AI Engineer: speaker page, Harrison Chase — checked 6 Nov 2026
  2. Fortune: LangChain is now a unicorn with a fresh $125 million in funding, October 2025 — checked 6 Nov 2026
  3. LangChain blog (Harrison Chase): Reflections on Three Years of Building LangChain, October 2025 — checked 6 Nov 2026
  4. LangChain: About — checked 6 Nov 2026
  5. LangChain blog (Harrison Chase): The rise of context engineering, June 2025 — checked 6 Nov 2026
  6. LangChain blog (Harrison Chase): Deep Agents, July 2025 — checked 6 Nov 2026
  7. LangChain blog (Harrison Chase): Your harness, your memory, April 2026 — checked 6 Nov 2026
Editorial profile compiled from public sources. Not an endorsement, and listing does not imply participation in this site. To correct or remove this profile, contact us.