big week at langchain, with a lot of launches:
1/ OpenWiki - auto generate a wiki of a github repo 2/ two different voice agent tutorials 3/ Harbor integration and tutorial for long running, stateful evals 4/ programatic subagents in deepagents (RLM like)
🧵with details
Users praise LangChain's stateful eval harness for agents because it enables reproducible testing that actually matters, along with appreciation for democratizing advanced AI tools via the tutorials and integrations.
No Digg Deeper questions have been answered for this story yet.
#6982/ Voice agent tutorials
voice agents are becoming increasingly popular, and we want to help you build, test, and observe them
first we worked with @pipecat_ai on an example of using a STT, TTS pipeline to turn any langgraph agent into a voice agent: https://youtu.be/i_Tf956Yh0U
Then we added an example of using deepagents as a subagent in native voice<>voice model: https://youtu.be/aV0ctbnDUCM
1/ OpenWiki
First, we wrote about why "wikis" are interesting as a form of memory: https://www.langchain.com/blog/wiki-memory
Cites examples from @FactoryAI, @cognition, @karpathy
Then @BraceSproul open sourced OpenWiki - a fully OSS way to create these wikis for your codebase https://github.com/langchain-ai/openwiki
4/ programatic subagents in deepagents
Deep Agents is our open source, model agnostic agent harness: https://github.com/langchain-ai/deepagents
We added the ability to programatically call subagents. We included examples of 6 common ways to do this: - Blog: https://www.langchain.com/blog/introducing-dynamic-subagents-in-deep-agents - Video: https://youtu.be/5AkdMangfNk
@sydneyrunkle showed how you can use this to do RLM like things: https://youtu.be/5_LLMZfKI6w
@huntlovell wrote about how we made the code interpreter we use for this secure: https://www.langchain.com/blog/running-untrusted-agent-code-without-a-sandbox
3/ Harbor integration
Harbor is a framework for running long running, stateful agent evals
We integrate with it in multiple different ways. We wrote a blog on those integrations: https://www.langchain.com/blog/unified-stack-for-evaluating-agents
We then made a video of how to use Harbor and LangSmith to evaluate deepagents: https://youtu.be/Rh6KWngr8T4
huge week of really exciting launches at langchain!
big week at langchain, with a lot of launches:
1/ OpenWiki - auto generate a wiki of a github repo 2/ two different voice agent tutorials 3/ Harbor integration and tutorial for long running, stateful evals 4/ programatic subagents in deepagents (RLM like)
🧵with details
1/ OpenWiki
First, we wrote about why "wikis" are interesting as a form of memory: https://www.langchain.com/blog/wiki-memory
Cites examples from @FactoryAI, @cognition, @karpathy
Then @BraceSproul open sourced OpenWiki - a fully OSS way to create these wikis for your codebase https://github.com/langchain-ai/openwiki
big week at langchain, with a lot of launches:
1/ OpenWiki - auto generate a wiki of a github repo 2/ two different voice agent tutorials 3/ Harbor integration and tutorial for long running, stateful evals 4/ programatic subagents in deepagents (RLM like)
🧵with details
3/ Harbor integration
Harbor is a framework for running long running, stateful agent evals
We integrate with it in multiple different ways. We wrote a blog on those integrations: https://www.langchain.com/blog/unified-stack-for-evaluating-agents
We then made a video of how to use Harbor and LangSmith to evaluate deepagents: https://youtu.be/Rh6KWngr8T4
2/ Voice agent tutorials
voice agents are becoming increasingly popular, and we want to help you build, test, and observe them
first we worked with @pipecat_ai on an example of using a STT, TTS pipeline to turn any langgraph agent into a voice agent: https://youtu.be/i_Tf956Yh0U
Then we added an example of using deepagents as a subagent in native voice<>voice model: https://youtu.be/aV0ctbnDUCM
so proud to be working with a bunch of people who are absolutely crushing it! check out all these new things if you haven’t had a chance to yet. huge week for us here @LangChain!!
big week at langchain, with a lot of launches:
1/ OpenWiki - auto generate a wiki of a github repo 2/ two different voice agent tutorials 3/ Harbor integration and tutorial for long running, stateful evals 4/ programatic subagents in deepagents (RLM like)
🧵with details
@hwchase17 this is awesome! really love how you guys are democratising the latest tech in ai
@hwchase17 the eval harness getting stateful is the part that actually matters here
@hwchase17 For me this is the part that actually matters. Reproducible evals for stateful agents. Cool demos are easy, clean environments and traces are the hard part.
