MLOps news and analysis on Web Pulse: 3397 curated articles tracking the latest MLOps developments, tools, and research, updated continuously from vetted sources.
Spotify's engineering team cut Claude Code token usage by 90% by routing routine I/O-heavy tasks to a smaller, cheaper model while reserving frontier models for complex reasoning, according to the company's engineering b…
A developer proposes a workflow to prevent coding agents from breaking hidden couplings during cleanup refactors. The method involves freezing a ledger of side effects and storing a SHA-256 hash before any structural cha…
A developer detailed how to cache Docker layers on Google Cloud Build using BuildKit's registry cache exporter, reporting a 94 percent cache hit rate across six production services. The approach pushes cache as OCI blobs…
AsyncGRPO, an asynchronous streaming pipeline for reinforcement learning post-training, eliminates the GPU idle bubbles that leave clusters idle 75% to 85% of total training time under synchronous Group Relative Policy O…
A benchmark of Qwen 3.8 27B across inference providers found Together AI FP8 TP2 delivered the highest aggregate output throughput at every concurrency level tested, reaching 2,701.47 tokens per second at concurrency 64 …
A developer argues that current agentic coding tools like Claude Code and Codex are not yet capable of reliably running 10 or more agents simultaneously, citing bottlenecks in isolation, compute, and merge management. Th…
LLM-as-judge architectures are shifting from offline evaluation into agent runtime control flow, letting applications make judgment calls on AI-generated work as it happens, according to a technical analysis. The approac…
CIQ shipped Fuzzball 4.2 on September 3, adding a native Model Context Protocol (MCP) server that lets AI agents inspect, submit, and monitor HPC jobs directly, plus a self-service workflow API enabling jobs to spawn oth…
A developer built TraceShrink, an OpenTelemetry Collector processor that replaces random tail sampling with cost-aware rules to cut LLM observability costs. The processor keeps traces based on cost thresholds, errors, an…
A developer argues that many AI system failures stem from weak security foundations rather than data quality or algorithmic issues. The post emphasizes the need for strong infrastructure boundaries, separation of duties …
A developer migrating an MCP server and two clients from FastMCP 3 to FastMCP 4 documented several breaking changes that compile but fail at runtime. Key issues include a pip in-place upgrade leaving an importable shell …
Black Box, under Chief Digital & AI Strategist Erik Jost, runs a production-grade agentic software development loop in which roughly eight AI agents—using Claude Code and OpenAI Codex—route work by task type and every pu…
A developer published an AI agent evaluation playbook, a repeatable test battery for vetting whether models running inside agent frameworks can be trusted with semi-sensitive content and real write access. The playbook w…
A developer explains why machine learning models that perform well during development often fail in production, citing data drift, improper data splitting, preprocessing inconsistencies, and concept drift as key causes. …
A developer has created a harness that versions golden behaviors to catch silent LLM regressions that survive snapshot tests. The method scores models with independent graders on obligations like grounded questions and b…
Antonio Lopes Correia, in part 8 of an experiment building an LLM-powered support agent, details a regression suite that evaluates agent behavior via scenario-based evaluators with per-property pass bars instead of stric…
A developer has released an open-source tool, llm-eval-harness, designed to catch regressions in LLM output by integrating evaluation into CI pipelines. The tool uses plain JSON for test data and supports matching rules …
Superlinked has released the Superlinked Inference Engine (SIE), an open-source server designed to consolidate the many small models used in AI agent workflows into a single deployment, addressing what the company calls …
A tutorial by Rachel Goldstein demonstrates how to regression-test LLM prompts using promptfoo 0.122.2 in GitHub Actions, catching prompt and model regressions on every pull request. The setup includes a support-ticket t…
A developer at xenition replaced five engineers with five AI agents for a week, reporting 23 pull requests opened, 14 merged, and 3 reverted. The experiment showed agents excelled at unambiguous, low-judgment tasks but s…