Langchain Community Kenya

Learn, build and share everything about LangChain and AI Agents. Weekly sessions, hands-on projects, Networking with AI enthusiasts

[it’kanyirijames @itskanyirijames]

Interviewing you before taking any action, so the agent doesn't make wrong decisions by guessing

@langchain @CopilotKit @claudeai

[Langchain Community Kenya @langchainke]

Join us this Wednesday, March 18th from 8:00 PM — 9:00 PM (EAT) as Roy gets into what it actually takes to build agents that think, remember, and act.

https://t.co/ZLzRjL5fk6

#LangChain #LangGraph #AIAgents #MachineLearning #NairobiTech #AIEngineering #BuildInPublic

[it’kanyirijames @itskanyirijames]

Interactive charts and diagrams

@CopilotKit @LangChain @claudeai

[it’kanyirijames @itskanyirijames]

Generative UI, directly in the chat.

@claudeai @CopilotKit @LangChain #agents

[Langchain Community Kenya @langchainke]

Join us for our next LangChain Community Kenya monthly meetup as we deep dive into:

​- What interrupts are
- How LangGraph implements them under the hood
- How to use them to add human oversight
- Real-world examples and demos

https://t.co/f5you5RXpt

[Langchain Community Kenya @langchainke]

Join our next LangChain Community Kenya meetup to explore what’s new in LangChain & LangGraph 1.0.

If you’re an AI builder, student, or enthusiast, come learn, share, and connect with others in the community.

#LangChain #LangGraph #AICommunity #KenyaAI

[swyx @swyx]

ok so let me explain why subagents kill long context

Like you can spend $500m building 100 million context models, and they would be 1) slow, 2) expensive to use, 3) have huge context rot. O(n) is the lower bound.

Cog's approach is something you learn in day 1 of @CS50 - divide and parallelize. Embeddings are too dumb, Agentic Search is too slow. So train limited-agency (max 4 turns), natively parallel tool calling (avg parallelism of 7-8, custom toolset) fast (2800tok/s) subagents to give the performance of Agentic Search under an acceptable "Flow Window" that feels immaterially slower than Embeddings.

The benefit of this is threefold:
- 8 ^ 4 toolcalls cover a very large code search space. can compound subagent calls if more needed.
- predictable cost & end to end latency
- subagent outputs "clean" contexts, free of context failure modes like context poisoning and context rot (h/t @dbreunig )

we originally called this Rapid Agentic Search, to contrast with RAG. but Fast Context rolls of the tongue better.

there's 2 other perspectives that are worthwhile, i'll go into below, but just go try it out. here it is on @karpathy's fastchat

[LangChain @LangChain]

We're hosting meetups in San Francisco, Boston, and NYC to celebrate LangChain's 3rd birthday and share some amazing news!

Come hang with our engineering team, get a first look at our newest launches, and connect with other builders shipping agents in production.

Plus: food, drinks, and good conversation about what's actually working (and what's hard) when building agents.

Whether you're in the Bay Area, or on the East Coast—pick your city and join us:

SF: https://t.co/HSJ3g3EDzF
Boston: https://t.co/gOj1ORNoLE
NYC: https://t.co/A2b9ZLEwn7

See you there 👋

[it’kanyirijames @itskanyirijames]

Had a great opportunity to speak about #Evals at #RenderCon2025 and how to use tools like #langsmith by @langchain to help when creating evaluations dataset and evaluators.

Big thanks to Rendercon Kenya for organizing the event.

#langchain #langgraph #langsmith #evals

[it’kanyirijames @itskanyirijames]

I will be speaking at AI Tinkerer Nairobi today alongside @gatere_mark and shadrack otieno

If you’ll be there, let’s connect.

[it’kanyirijames @itskanyirijames]

Day 10/20 of the #AIAgentADay Challenge

Today I added a supervisor agent to improve research quality in multi-agent workflows.

Why does this matter? Let me explain 🧵

[it’kanyirijames @itskanyirijames]

Day 4 of the #AIAgentADay Challenge

I built an agent to help draft a research paper by writing the sections in parallel

Why this matters: Speed
Pattern used: Orchestrator-Worker

Check it out: https://t.co/aAtN61L2Da

4 down, 16 more to go!

[Langchain Community Kenya @langchainke]

Join us for “The Fundamentals of Vibe Coding: Principles, Types, and Applications”

�� Date: Sep 10
🕖 Time: 7:00PM – 8:00PM
🌍 Where: Online

It’s going to be an engaging and insightful session you don’t want to miss!

RSVP:https://t.co/nMNQFf6QjZ

[LangChain @LangChain]

🤖🔌 DeepMCPAgent

A powerful tool for dynamic MCP tool discovery and agent development. Built with LangChain and LangGraph, it streamlines integration over HTTP/SSE while supporting major LLMs.

Check it out! 🚀
https://t.co/4CGOb04Ycu

[LangChain @LangChain]

‼️LangChain & LangGraph 1.0alpha releases

Today we are announcing alpha releases of v1.0 for langgraph and langchain, in both Python and JS.

🕸️LangGraph is a low-level agent orchestration framework, giving developers durable execution and fine-grained control to run complex agentic systems in production.
🦜LangChain helps developers ship AI features fast with standardized model abstractions and prebuilt agent patterns, making it easy to build complex applications without vendor lock-in.

We are working towards an official 1.0 release in late October - please give us any feedback!

TL,DR:
1. LangGraph is largely the same as before, no breaking changes. We’ve heard good feedback about LangGraph and are excited to move to 1.0
2. New standard content blocks on messages in LangChain Core. Message formats are evolving, and so is LangChain. Backwards compatible
3. LangChain itself - high level agents and chains - is greatly changed. You should think of LangChain 1.0 as a new package - focused around a central agent abstraction, built on top of LangGraph
4. New docs site!

You can find a discussion topic here: https://t.co/wojyNFwgNu

Please give us feedback!

[Langchain Community Kenya @langchainke]

We’re excited to welcome @JimmyMuthoni as the LangChain Community Kenya Campus Ambassador for @DeKUTkenya

Looking forward to the great impact he will create in empowering students to build with @langchain

[Langchain Community Kenya @langchainke]

We’re excited to welcome @JimmyMuthoni as the LangChain Community Kenya Campus Ambassador for @DeKUTkenya

Looking forward to the great impact he will create in empowering students to build with @langchain

[Langchain Community Kenya @langchainke]

Agentic apps are powerful but unpredictable small changes in prompts, models, or inputs can lead to very different results. That’s why evaluation is key.

Join us to learn how to evaluate your agents led by @itskanyirijames
RSVP: https://t.co/kOYC3PFN0D

#LangChain #LangSmith