LangChain

Headed to #SFTechWeek? We’re partnering with Modal + Cogent to host a live, team-based AI Heist challenge. You'll work in a fictional, isolated cloud environment, direct an agent through staged security challenges to find the hidden proof token, and compete on a live leaderboard for prizes. Capacity is limited. RSVP today! https://lnkd.in/guRqR6n7

Next stop on the LangSmith Roadshow? Los Angeles! Join us on October 1st for a half-day of keynotes, deep dives, and magic. https://lnkd.in/gQK99YKs ✅ Learn about the state of agent development ✅ Go hands-on with LangSmith Engine and Deep Agents ✅ Hear how ServiceTitan is using LangSmith in production as well as the lessons they’ve learned along the way 🪄 We’ll wrap up with networking and an exclusive live magic show just for our group Spaces are limited, so RSVP today!

In Managed Deep Agents 0.8, we introduced HTTP channels. Agents can now take requests from any webhook, not just Slack. Agents can now live inside customer portals, support queues, order systems, or anywhere a webhook can reach. Nathan Drezner built one that runs entirely over text message. Text it a receipt, it triages the expense and logs it straight to memory. Watch his demo, and see what else is new in Managed Deep Agents at https://lnkd.in/gmigVEAh

What does it take to move from isolated agent experiments to production AI at enterprise scale? Our latest guide brings together lessons from Schneider Electric, Vodafone, and monday.com across: - Shared agent platforms and LLMOps - Production observability and evaluation - Multi-agent architectures - Clearer boundaries between agents, tools, and environments - Security, governance, and data residency - Scaling adoption across teams and business units Download Scaling Agents in Europe & the Middle East to explore how leading organizations are building the foundation for production agents at scale: https://lnkd.in/gqTTahry

Igor Nikolaienko, AI Architect at Deutsche Post und DHL is taking the stage at Interrupt, The Agent Conference by LangChain. Catch Igor’s talk + more: https://lnkd.in/gUv4ce-j

A customer says the agent got it wrong. LangSmith is how Clay finds out why.

Debugging agents often starts with a simple question: what did the agent actually do? Finding the answer in short, single-turn workflows has usually been simple. For longer-running agents, even a single session has required sifting through user turns, tool calls, retries, and subagent handoffs. At Interrupt, we launched Trajectories in LangSmith to make that trace data easy to navigate: ✅ Every unit of work an agent does gets recorded as a run, and a trajectory strings the messages and actions across all of them, main agent and subagents included, into the order they happened ✅ This provides the readable path the agent took, with the full trace still there when a step needs a closer look Read on to see how teams can find root causes faster, get expert feedback with less friction, and continuously improve with production behavior data. https://lnkd.in/g_XM8V9Z

Meet the new terminal interface in Managed Deep Agents.

Agents can help life sciences teams work across scientific literature, clinical trial data, code, and analysis—but building a research agent that can reliably handle complex scientific workflows requires more than connecting a model to tools. Join us for the Deep Life Sci Workshop: Agent Engineering for Life Sciences, where we’ll use Deep Life Sci to explore the agent engineering patterns behind more accurate, observable, and reliable research workflows. We’re hosting sessions for both NAMER and EMEA audiences. 🌎 NAMER: https://lnkd.in/gtcd9Gmv 🌍 EMEA: https://lnkd.in/gdSeyp25