How Orange Business Built Live Intelligence with LangChain and LangSmith
Orange Business used LangChain and LangSmith to build Live Intelligence, a platform for agents with governance, evals, observability, and production scale.
Orange Business used LangChain and LangSmith to build Live Intelligence, a platform for agents with governance, evals, observability, and production scale.
The Deep Agents SDK manages context for long-running AI tasks by preventing context rot through offloading, summarization, and filesystem abstraction.
LangChain used the Deep Agents CLI to automate migrating code samples into testable, CI-ready examples, addressing the problem of stale code snippets in documentation.
AI Search in LangSmith allows querying traces using natural language by describing the behavior you're looking for to retrieve relevant trace data.
Use ScaleDown's constrained decoding models with LangChain and Deep Agents to summarize history, compress tool results, and extract data faster and cheaper.
GitHub administrators can now track the effectiveness of AI Scan for pull requests through new metrics in the security overview.
Enterprise administrators can now use telemetry in managed settings to attribute GitHub Copilot activity to a user without collecting their prompts, responses, or tool content.
The langgraph-checkpoint update to version 4.3.0 includes fixes for InMemorySaver, namespace label validation, and replaying delta writes.
GitHub Copilot for JetBrains introduces new controls for default models and the MCP server. The update simplifies diagnostics, improves chat navigation, and account management.
This week's GitHub Copilot updates improve cross-account and cross-environment usability with more control over agent access.
OpenAI Codex CLI 0.162.0 adds tools for managing Git worktrees, pinning tasks, navigating transcripts, and clickable URLs.
Google is launching the Developer Knowledge API and MCP Server in public preview. This toolset enables AI assistants and agent platforms to search and retrieve up-to-date Google documentation.
AWS DevOps Agent can now turn investigation summaries into pre-validated fixes that an engineer can approve with a single action, using AWS Lambda Durable Functions, Amazon EventBridge, and Amazon Bedrock.
GitHub Copilot now offers local sandboxing in GitHub Copilot CLI, the GitHub Copilot app, and VS Code sessions. This provides a secure environment for developers.
GitHub Copilot CLI now allows developers to select local models without leaving their workflow. The `/model` command in version 1.0.94-0 discovers supported models.
Google AI has launched Playground, a new experimental gaming platform enabling users to create, play, and share custom games. The platform is now available for public access.
GitHub has fixed the cause of declining agent activity and agent lines of code in Copilot usage metrics. Updating your IDE will restore these metrics.
Release 0.161.0 has been published for OpenAI Codex CLI rust.
Vercel AI Gateway can now escalate a decision request to a fallback model when the primary model's answer trips a confidence condition. This applies to Choice and Score questions, with Boolean questions escalating based on a probability range.
AWS provides tools for customers to align with international standards for responsible AI governance, including AI system impact assessments to improve enterprise risk management.
Organization and enterprise administrators can now view the enablement status of AI Scan for pull requests within GitHub's security overview coverage view.
AWS introduced automated, auditable cross-account promotion of Amazon Quick resources using an MCP server on Amazon Bedrock AgentCore.
Amazon Bedrock Knowledge Base with LangChain enables building RAG applications with agentic retrieval for multi-part questions.
Amazon Quick offers two methods for downgrading user roles: a manual delete-and-recreate approach and an AWS CLI step-down sequence that safely downgrades roles while preserving asset ownership.
OpenAI introduces a new visual ad format in ChatGPT and expands measurement tools, attribution partnerships, and brand suitability for advertisers.
GitHub Copilot now supports APIs for code review requests and setting effort levels, with 'Balanced' as the default.
LangChain presented a new method for building a model router, which reduced the median cost per coding task by 64% with no measurable drop in quality.
6 to build technology and redesign workflows, cutting trade validation from 30 minutes to under 4.
NVIDIA TensorRT RTX Samples show how to add AI models to local applications using a portable model format, reliable runtime, and acceleration.
Albertsons Cos. is using ChatGPT Enterprise and the OpenAI API to help teams work faster and make grocery shopping easier for millions of customers.