MLOps news and analysis on Web Pulse: 3301 curated articles tracking the latest MLOps developments, tools, and research, updated continuously from vetted sources.
A developer behind Halfcycle, a five-layer build method that runs inside Claude Code, documented nine blockers found across five review rounds on a small document-restore feature before any code was written. Four of the …
A developer cut token costs by roughly 60% by replacing per-agent model bindings with per-call capability-based routing, sending only the calls that need frontier-class reasoning to expensive models. In a 1,000-review te…
A second-generation Spanish family business servicing and repairing car-wash equipment has built its own ERP in production and a set of AI agents on Claude Code and an MCP server, despite having no programmers on staff. …
Arize AI aligned the automated evaluators for its Alyx AI assistant in Arize AX by sampling production sessions into a labeling queue and building a Custom View that places the full session transcript, tool sequence, err…
A developer benchmarked five ways of running the same AI classification job on fake customer support chats — four batch API services (OpenAI, Gemini, Claude with thinking on and off) plus a self-hosted model — and found …
A developer recommends that API-only AI tutoring startups adopt a lean observability stack: exception capture with searchable, stable failure groups plus a small set of cost-allocation fields attached to every event, whi…
TypeSafe introduced Jev, a specialized model for bounded decisions between an agent's reasoning and its actions, and Cloudflare and AWS followed last week with Clef/Clef-flash and Strands Decider 2B, respectively, as ent…
A deep-dive walkthrough documents ML.NET, Microsoft's open-source, cross-platform machine learning framework that lets .NET developers build, train, and run models entirely in C# or F# without leaving the .NET ecosystem.…
A developer behind the MCP security scanner WARDEN benchmarked it against two open-source scanners, mcp-audit and mcp-shield, on 485 poisoned tool definitions extracted from the MCPTox dataset and on 986 public servers f…
TaskHandoff launched as a self-hosted control plane for running, managing, and collaborating with containerized AI agents across local and remote machines, with Docker as its primary isolated runtime. The open-source pro…
MongoDB launched Atlas Agent Engine, an agent runtime and memory platform now in public preview on AWS, positioning its database as the agent control plane. MongoDB Chief Product Officer for AI and Emerging Products Pabl…
Ai2's AI Infrastructure team replaced its priority-based GPU cluster scheduler with a system combining GPU time budgets, hierarchical fair-share allocation, and a time-slicing contract, shifting GPU-time disputes from ca…
BFI Finance moved roughly 170 developers and 180+ repositories from SonarQube to Codacy's seat-based static analysis platform, cutting direct static analysis costs by 18 to 20%, according to the Indonesian financial serv…
A developer built Oraviz-MCP, a minimal, read-only Model Context Protocol server for Oracle AI Database that exposes seven tools for running queries, profiling tables and generating charts, as a proof of concept for lett…
A developer published a practical security checklist for MCP servers, arguing that because tools touch filesystems, databases, package managers and deployment APIs, the server itself becomes the authorization boundary. T…
A developer ran a prompt in Claude Code asking which tools the agent was missing, and the model instead returned a prioritized audit of the developer's broken configuration, flagging a dead MCP server, a CLAUDE.md rule i…
Narwhal, an open-source adaptive disaggregated LLM inference framework, hot-swaps prefill and decode roles across a fixed GPU fleet in seconds without reloading model weights, scaling from a single GPU to multi-node depl…
A green CI result does not prove architectural fit for AI-generated Terraform, according to a practical review checklist published by CloudGeni that calls for plan evidence and post-change verification in pull requests. …
A developer outlined a four-engine architecture for deploying AI on physical assets such as manufacturing plants, logistics centers, and autonomous machines, arguing that standard cloud API, microservice, and caching pat…
Metasploit's latest release adds 12 new modules, including four targeting LLMs and AI infrastructure: a scanner for CVE-2026-22778 in vLLM's multimodal endpoint, an unauthenticated RCE in Langflow versions 1.7.3 and belo…