Native Integrations for AI Agents: LangChain, LangGraph, LangSmith

This title was summarized by AI from the post below.

911,693 followers

Going from prototype to production with AI agents means managing retrieval, memory, and operational data as separate problems. Now, native integrations across LangChain, LangGraph, and LangSmith combine operational data, vector search, and persistent state all in one place, without standing up parallel infrastructure. Read the full announcement 👇

LangChain

508,016 followers

3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww

2 Comments

Evan Kirstel 2w

Consolidating retrieval, memory, and operational data into a single platform is a game-changer for AI agent development. The LangChain/LangGraph/LangSmith integration eliminates the infrastructure sprawl that slows most teams going from prototype to production. MongoDB is becoming the de facto data layer for the agentic AI stack. Would love to explore this on my podcast @evankirstel — check out techimpact.tv!

IndexScore 3w

Spot on—productionizing agents is really about managing state, retrieval, and ops together. Unifying them reduces friction fast. At IndexScore, we see similar gains when data and context stay tightly connected.

See more comments

To view or add a comment, sign in

More Relevant Posts

LangChain

508,016 followers
3w Edited
Report this post
📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
24 Comments
Like Comment
To view or add a comment, sign in
Karan Singh
3w Edited
Report this post
We are deepening our MongoDB partnership to make it easy for enterprise customers to build agents securely on top of the data layer they already use. Read more about how you can leverage MongoDB and LangSmith! 🤝🏼 Grateful to Harrison Chase and Chirantan "CJ" Desai for their leadership in bringing this to life. And, special thanks to the great team behind the scenes: Gregory Maxson, Julia Schottenstein, Aki Dewakar, Harshada Kamalapurkar, Zara Pasalar, Adrienne Cohen, William Fu-Hinthorn, Rebecca Weng and Jess Ou!
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
1 Comment
Like Comment
To view or add a comment, sign in
Harshada Kamalapurkar
3w
Report this post
A big step forward for teams looking to move from AI experimentation to real-world impact. By bringing together LangChain’s leading framework for building AI agents with MongoDB's developer data platform, it becomes easier to build, scale, and productionize GenAI applications on a unified stack. Looking forward to the innovation this drives! 🚀 ➡️ Read the announcement: https://lnkd.in/euhrUfN6
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
2 Comments
Like Comment
To view or add a comment, sign in
Logan Frost
3w
Report this post
Last week we had the opportunity to hear directly from LangChain's CEO on how powerful MongoDB's document model will be on the next wave of AI native applications. In his words: "JSON is the closest thing to natural language for LLM's to consume" SO excited for MongoDB to continue to be recognized as the standard for building AI applications. #LLMsLoveJSON
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in
Andy P.
3w
Report this post
Excited to see MongoDB and LangChain make it easier for teams to build production AI agents on the data foundation they already trust.
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in
Ben Karciauskas
3w
Report this post
As anyone following along knows, I built a personal AI assistant I call Clawdia. She runs on OpenClaw with Claude as the brain. But the bit that ties it all together? MongoDB. Conversation history, customer interactions, a knowledge vault with vector search, persistent memory across sessions — all in Mongo. When I need to find something from three weeks ago, it searches semantically. When it crashes, it picks up where it left off. When I ask "what happened in last week's customer call?", it queries my operational data and gives me an answer. I didn't plan this architecture. I just kept reaching for MongoDB because the data was already there. Why would I bolt on a separate vector database, a separate state store, and a separate analytics layer when one system already does all three? Turns out LangChain and MongoDB just formalised exactly this pattern. Atlas as the complete AI agent backend: vector search, persistent agent memory, text-to-MQL for querying operational data in plain English, and full observability through LangSmith. The bit that matters for enterprise teams: you probably already run MongoDB. Your AI agents don't need a new data layer. They need access to the one you've got.
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in
Chentel Johnson
3w
Report this post
LangChain x MongoDB 🤝 “Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end-to-end, from retrieval calls to routing decisions to state transitions.” MongoDB = built for the AI-Era 🌟
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in
ENCS

13,485 followers
3w
Report this post
🚀🧠 MongoDB nails the core challenge of production AI agents! Managing retrieval, memory, and operational data as separate silos is a recipe for friction — the teams that win will be those who unify these layers into a coherent, scalable architecture. This insight from MongoDB is a must-read for every AI engineer and tech leader building agent-based systems today. The future of AI is not just models — it's the data infrastructure powering them! 💡 👉 Follow for more AI & tech insights: https://lnkd.in/enyqudEU | https://lnkd.in/en7PUPsX #MongoDB #AIAgents #GenerativeAI #LLM #AIInfrastructure #VectorDatabase #RAG #MachineLearning #DataEngineering #CloudComputing #TechLeadership #ArtificialIntelligence #AIProduction #AgenticAI #Innovation
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
1 Comment
Like Comment
To view or add a comment, sign in
Bryan Augustin
3w
Report this post
If your team is building AI agents with LangChain, this is worth 2 minutes of your time... MongoDB and LangChain have partnered to give you a complete path from prototype to production on the database you already use. No bolt-ons, no new infrastructure to manage If you're evaluating where your #agentmemory and state should live, the answer just got a lot clearer
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in
Halim Khalifa
3w
Report this post
Moving AI projects from a cool prototype to actual production usually means stitching together a bunch of different, complex tools. Being able to handle vector searches and agent memory natively inside MongoDB Atlas is a total game-changer because it eliminates the need to stand up and pay for parallel infrastructure. It is really interesting to see trusted, traditional databases evolve to handle these complex AI workloads natively, making life a lot easier for developers. #AI #AIStack #VectorSearch #LangChain #MongoDB #SoftwareEngineering #TechInnovation
LangChain

508,016 followers
3w Edited

📣 Announcing our partnership with MongoDB: The AI Stack that runs on the database you already trust We've partnered with MongoDB so teams can go from prototype to production leveraging existing Atlas deployment, without standing up parallel infrastructure. Atlas Vector Search plugs directly into langchain as a drop-in retriever for semantic, hybrid, and GraphRAG queries. The MongoDB Checkpointer for LangSmith Deployment persists agent state (multi-turn memory, human-in-the-loop workflows, audit trails) directly in Atlas. Text-to-MQL lets agents query operational data in plain English. And LangSmith traces everything end to end, from retrieval calls to routing decisions to state transitions. ➡️ Read the announcement: https://lnkd.in/euhrUfN6 ➡️ Check out the MongoDB docs: https://lnkd.in/efnMyncJ ➡️ Check out our docs: https://lnkd.in/eewuc9Ww
Like Comment
To view or add a comment, sign in