LangChain

Evidence (14)

• 5InfoQ·10/5/2026framework_update
  Java roundup: JobRunr 9, OpenXava 8, and new updates across the Java stack

  This Java roundup highlights GA releases of JobRunr 9.0 and OpenXava 8.0, along with point releases for Quarkus, Micronaut, LangChain4j, and Eclipse JNoSQL. It also notes the introduction of ADK for Java and Kotlin, plus Lathe, a new Java language server.

• 4InfoQ·9/25/2026research
  Exploring the Agent Harness: Two Ways to Build One

  This article explains what an agent harness is and compares two implementations using a finance assistant example: AWS AgentCore Harness and LangChain with Envoy AI Gateway. It focuses on how the two approaches handle tools, memory, model access, cost control, observability, and the trade-offs between portability and operational ownership.

• 7The New Stack·9/10/2026open_source
  AWS open-sources Pizza Bot for managing background AI agents

  AWS has released Pizza Bot, an open-source desktop app that gives developers an email-style inbox for monitoring and triaging background AI agents. The project is self-hosted and uses LangGraph-based persistence so agent work can pause, resume, and surface updates even when the user is away.

• 4InfoQ·9/7/2026framework_update
  Java roundup highlights TornadoVM 6.0 and multiple point releases

  This Java roundup highlights the general availability of TornadoVM 6.0 alongside point releases for JReleaser, LangChain4j, Java Operator SDK, JHipster, Kotlin Toolchain, and Yupiik Fusion. It also notes maintenance releases for Micronaut and the GraalVM Development Kit.

• 7The New Stack·8/20/2026product_launch
  Slack simplifies installing third-party AI agents

  Slack launched Add to Slack, a new installation flow that lets users add agents built on partner platforms without creating a custom Slack integration. The runtime stays outside Slack, while OAuth, app configuration, and permission scoping are handled by the partner tool; installed agents also inherit workspace access controls and appear in Slack’s app browser.

• 5The New Stack·8/20/2026research
  Building Token-Efficient Multi-Agent Systems

  The article argues that the main cost in production AI agent systems is often not the model itself, but redundant retrievals, repeated prompts, duplicate tool calls, and oversized context windows. It outlines a token-efficient architecture using routing, semantic caching, context budgeting, adaptive retrieval, and model selection to reduce latency and cloud spend without lowering output quality.

• 8The New Stack·8/18/2026research
  Agentic AI faces a latency wall that more compute cannot fix

  Akamai’s State of AI Inference 2026 report says half of enterprise AI deployments are missing their latency targets at peak load, even as most critical use cases now require sub-500ms response times. The article argues that agentic workflows create many sequential hops across networks and CPUs, so adding more GPUs alone does not solve end-to-end latency in production.

• 7The New Stack·8/7/2026framework_update
  Coinbase, Shopify and Ramp build in-house coding agents while still using Anthropic

  Coinbase, Shopify, and Ramp have each developed internal coding agents for software engineering workflows, but they continue to rely on commercial tools and frontier models from Anthropic, OpenAI, and Google. The article argues that enterprises are increasingly owning the agent harness — the execution layer for context, permissions, orchestration, tool use, and verification — rather than the model itself.

• 4InfoQ·7/24/2026research
  Self-Building Agent Experiment with LangChain4j

  The article describes an experiment in which a code assistant used LangChain4j documentation to design an agentic system. The assistant built a framework that could write, test, and debug code autonomously, with supervisor and workflow architectures showing different trade-offs between flexibility and debugging speed.

• 6The New Stack·7/23/2026research
  Nvidia Says Local and Frontier Models Will Work Together

  Nvidia executive Joey Conway said organizations will increasingly use local/open models alongside frontier models, with routing systems choosing the best model for each task. He also pointed to enterprise control, lower latency, and lower cost as reasons to run adapted open models near the data, citing Nvidia hardware and software used to serve and orchestrate these workloads.

• 5The New Stack·7/21/2026research
  High-reasoning AI coding emerges as the next frontier

  The article argues that single-pass AI coding is still useful for simple tasks, but higher-reasoning workflows are becoming more important for complex, security-critical, or architectural work. Industry voices suggest teams should route tasks to the right model rather than treating deeper reasoning as a universal upgrade, while noting that high-reasoning systems can be more of a black box.

• 6InfoQ·7/20/2026framework_update
  Java roundup highlights Value Objects preview, WildFly 41 GA, and AI tooling updates

  This Java news roundup covers a preview reintroduction of Value Objects, the general availability of WildFly 41, and the July 2026 release of Open Liberty 26.0.0.7. It also notes point and maintenance updates across TornadoVM, Apache TomEE, Java Operator SDK, LangChain4j, Micronaut, plus the new Quarkus Shim extension and Oracle AI Agent Studio for Fusion Applications.

• 4InfoQ·6/29/2026framework_update
  Java Roundup: Hardwood and Endive GA, Payara Update, and WildFly Beta

  This Java roundup covers GA releases of Hardwood 1.0 and Endive 1.0, along with the June 2026 release of Azul Payara. It also notes point releases for Quarkus and LangChain4j, the first beta of WildFly 41, and the introduction of Eliya JDK and the Open Source Sustainability Initiative (OSSI).

• 6Hugging Face Blog·6/23/2026open_source
  IBM’s CUGA shows a lightweight harness for agentic apps

  IBM’s Hugging Face Blog post introduces CUGA (Configurable Generalist Agent), an open-source agent harness that handles planning, execution, tool calls, and state management so developers can focus on tools and prompts. The company also published two dozen single-file FastAPI examples to demonstrate how the same agent can be run in governed or sovereign production setups without redesigning the app.