www.infoq.com

AIAI-generated summary Walmart Global Tech deployed multi-agent LLM workflows across 4,700 stores for nightly batch inventory forecasting. Aditya Mulik explains that shifting from PTUs to pay-as-you-go reduced cloud inference costs from over 100k dollars to 10k dollars per month.

AIAI-generated summary InfoQ reports on reliability layers tested with language models on the CSIRO Adverse Drug Event Corpus to measure accuracy, yield, and verification limits. The evaluation found that the tested reliability layers sorted answers by trustworthiness but corrected almost none of them.

AIAI-generated summary Google released Gemini 4 Argon as a general-purpose physical-reasoning model intended for robot control, simulation, video labeling, and policy improvement. Google reports that Gemini 4 Argon achieves an 87 percent score on the Embodied Reasoning Question Answering benchmark and is available to subscribers and through the Gemini API.

AIAI-generated summary InfoQ reports on optimizing distributed AI inference using llm-d and vLLM on Kubernetes to manage cache locality and batch inference. Edoardo Vacchi discusses running Java workloads on GPUs using TornadoVM and the evolving landscape of WebAssembly on the backend.

AIAI-generated summary Amazon introduced TorchNeuron, a native PyTorch backend that allows developers to run existing code on AWS Trainium chips without modifications. Amazon says a single configuration change switches hardware, and the backend supports distributed training and PyTorch parallelism APIs.

AIAI-generated summary Pinterest replaced its journey clustering pipeline with a fine-tuned 4B Qwen3 LLM to infer user journeys from activity histories. Pinterest reports that the new model delivers improved engagement, including higher email click-through rates and increased push opens.

AIAI-generated summary Grab reports that the adoption of AI coding assistants has accelerated development and increased the volume of code changes per release, bringing more opportunities for defects and crashes. To address these challenges, Grab implemented a layered defense strategy involving static code analyzers, automated testing, staged rollouts, and real-time crash monitoring to maintain mobile app stability.

AIAI-generated summary Tiger Data has released the open source PostgreSQL extension pg_vectorsearch, featuring a PRISM index designed to reduce memory requirements and index build times. Tiger Data says the alpha release enables faster vector search directly in PostgreSQL, though incremental rebalancing is not yet fully implemented.

AIAI-generated summary Google created agents that scrubbed codebases to find bug references marked as fixed or obsolete for cleanup. Google says the todo cleaner removed thousands of obsolete references over six months with high acceptance from human reviewers.

AIAI-generated summary GitLab released GitLab 19.5 on October 15, integrating agentic development and governance into the DevOps platform. The release introduces a generally available GitLab MCP server, security controls for external agents, and the general availability of GitLab Secrets Manager.

AIAI-generated summary AWS recently made Continuum generally available for code vulnerabilities in US East and US West after an earlier gated preview. AWS prices the service with a context fee based on lines of code protected per month and an inference fee for findings worked.

AIAI-generated summary Advait Patel shares a playbook for securing AI agent systems in production that is available through InfoQ. The playbook covers moving defense boundaries to the tool layer and implementing kill switches to protect critical data.

AIAI-generated summary Uber has rolled out an MCP Gateway to govern access to more than 800 servers and 5,000 AI agent tools across the company. The platform provides a common layer for discovery, protocol translation, authorization, and execution to standardize how internal services are accessed.

AIAI-generated summary Amazon CEO Andy Jassy announced that AWS capacity is largely sold out through 2027, with portions of 2028 already allocated. The company continues to reserve headroom for On Demand, EC2 and Bedrock despite the high demand for committed capacity.

AIAI-generated summary Linearbytes released Bug0 QA Agent, a source-available Playwright library designed to automate regression testing using cached Playwright replays. The framework utilizes consensus from multiple language models and evaluations from vision models to handle assertion verification during testing.

AIAI-generated summary Netflix makes available its GPU reliability platform, which uses layered health checks, automated remediation, and fault-tolerant training to keep jobs running despite hardware failures. Netflix reports more than 99.95% GPU-node availability and an increase in job success from approximately 98% to more than 99.5%.

AIAI-generated summary Cloudflare previews a Web Search API through AI Gateway to give applications access to live web results from Ceramic.ai, Exa, and Linkup. Cloudflare states that the API allows AI applications to retrieve current web content and provide the results as context to language models during inference.

AIAI-generated summary Anthropic says Code Review for Claude Code is now available on Team and Enterprise plans. The security review runs as an open-source GitHub Action that is not hardened against prompt injection attacks.