LangSmith: The platform to run, improve, and govern agents @LangChain
LangSmith: The platform to run, improve, and govern agents  @LangChain
Uploaded November 2026 | Updated November 2026, 3 hours ago
LangChain has raised $300 million in a Series C at a $3 billion valuation, and we're launching the new LangSmith, which puts what we've learned from millions of developers building agents into one platform. LangSmith processes billions of traces a month for more than 7,000 customers.

The agents that change how a business runs get better because their teams keep iterating on them, and that work is spread across tools that weren't built for agents. This video walks through how LangSmith handles the loop in one place. You build an agent in natural language or code and deploy it in one click. Once it's running, you can trace every step it takes, and evals test each change before it ships. Workday used LangSmith to cut the time to deploy a new agent from five months to under a day.

LangSmith Engine monitors your agents around the clock for failures and writes the fixes and evals, so your part is reviewing and merging them. Teams using Engine improve their agents five times faster, and it's free for all new LangSmith Plus subscribers through the end of the year.

Chapters:
0:00 LangChain's Series C and the new LangSmith
0:15 What sets apart the agents that change a business
0:45 How to build and deploy an agent in LangSmith
1:03 How to trace and evaluate your agents
1:16 How LangSmith Engine finds and fixes agent failures
1:28 Governing every agent run
1:38 LangSmith Engine free for new Plus subscribers

Resources:
→ Start building with LangSmith: langsmith.com
→ LangSmith Engine: langchain.com/langsmith/engine
→ LangSmith Engine docs: docs.langchain.com/langsmith/engine
→ Pricing and Plus plan: langchain.com/pricing
LangSmith: The platform to run, improve, and govern agentsBuilding a Production Agent Eval Pipeline: Harbor + LangSmith + OpenAI SDKOpenWiki 0.2 is adopting the OKF specCreate an agent that can browse the web with Managed Deep Agents and Browserbases StagehandBuilding Semiconductor Agents at Skyworks with LangGraph + LangSmithWhy Distillation Might Be Impossible to StopThe LangSmith Feature Built for Multi-Step AgentsThe Future of AI Agents with Andrew Ng | Interrupt 26Build a social media agent with Managed Deep AgentsGetting Evals Right for LLM Applications | Interrupt 26CLI vs MCP: Which Should You Use for Building AI Agents? #LearnOnShortsDeep Agents in 1 sentence
LangChain |

LangSmith: The platform to run, improve, and govern agents

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER