News Hub

LangChain offers managed deep agents with durable execution and observability in private beta.

LangSmith LLM Gateway provides runtime governance with spend limits and PII redaction.

Managed Deep Agents now publicly available with durable execution and production infrastructure.

Deep Agents, LangChain, and LangGraph offer distinct approaches for building agents with different use cases.

LangChain built an autonomous SRE agent for Kubernetes using Deep Agents with human approval and LangSmith tracing.

LangChain documented production lessons from CX agents deployed at Lyft, Vodafone, and LATAM Airlines.

LangSmith provides methods to evaluate voice agents using traces, code evaluators, LLM judges, and human review across execution, outcomes, and caller experience.

Stripe built Kai, a company-wide AI agent using LangChain and LangGraph, reaching 5,000 users in approximately four weeks.

LangChain created ReviewBench, a benchmark for evaluating code review agents against real pull request feedback from trusted reviewers.

LangChain explains why coding agent costs escalate and provides methods to trace, compare, and control spending across multiple tools.

LangChain built a data stack using Hex, dbt, and semantic models to create a trusted data agent scaling self-service analysis 40x.

Companies must own their agent systems, governance, context, and feedback loops to achieve lasting AI competitive advantage.

LangChain releases Align Evals to calibrate LLM evaluators to match human preferences.

Similarweb uses LangSmith to evaluate agent research reports with rubrics, faithfulness checks, and traces.

Deep Agents v0.7 reduces base input tokens by 65 percent at comparable performance.

SmithDB implements full-text search with 400ms median latency over nested JSON in object storage.

LangChain revamped benchmarking for Deep Agents across coding, conversation, and retrieval tasks.

LangChain released NemoClaw Deep Agents blueprint, LangSmith Sandboxes, and other platform updates.

LangChain introduced Eval Engineering Skill to generate evaluations from repository context and execution traces.

Schneider Electric built enterprise LLMOps foundations using LangSmith for observability and evaluation.