MLOps news and analysis on Web Pulse: 3397 curated articles tracking the latest MLOps developments, tools, and research, updated continuously from vetted sources.
Mohammed Nihal03 released AI Agent Observatory, an open-source local dashboard that analyzes Claude Code sessions to score agent health from 0 to 100 and flag degrading behavior, with the tool catching its own author edi…
A Neo4j semantic layer built with Neocarta cut Text2SQL token costs by up to 81% on a 278-table BigQuery Census ACS catalog and up to 54% on a 264-table Databricks catalog, according to a Neo4j solutions engineer's bench…
An engineer at MyVitals detailed a month-long incident where an AI pipeline silently failed 100% of the time while logs reported success. A misconfigured environment variable caused every metric-matching LLM call to retu…
A developer argues that model-centric observability is insufficient for debugging retrieval-augmented generation (RAG) systems, because a green LLM trace can still hide a wrong answer caused by upstream retrieval failure…
Prefix caching can push LLM prefill from compute-bound to memory-bandwidth-bound once enough of the prefill is loading cached KV rather than computing it, according to a preprint by Meng, Lee and Wang, "Understanding Bot…
Spotify Engineering cut Claude Code token usage by roughly 90% on a Java monorepo by routing bulk file reading and boilerplate code generation to a cheap model such as Gemini 2.5 Flash while keeping reasoning on frontier…
A new architectural guide proposes a concrete cancellation contract for stopping AI agents on on-premises infrastructure, separating deadline enforcement from cancellation requests and tracking distinct states such as ru…
Monitoring of the mistralai/mistral-nemo endpoint on deepinfra/fp8 was retired on 2026-09-06 after the B3IT border-input detector could not find enough border inputs, and the endpoint returns no logprobs, making logprob …
Anshvardhan Shetty, a first-year EECS undergraduate at Imperial College London and research intern with Professor Juncheng Yang at Harvard SEAS, presents a study of KV cache admission on high-bandwidth flash using a 222-…
AC2, an end-to-end infrastructure platform for training and inferencing open weight models, has launched with SDK, runtime, and tracing packages that support coding agents like Claude Code, Cursor, and Codex. The platfor…
Overmind Lab details a four-stage pipeline for converting production agent traces into fine-tuning datasets, emphasizing standardized capture via OpenTelemetry GenAI conventions and strategies like failure-weighted sampl…
A developer at Cognisant LLC detailed a workflow for running six or more Claude Code sessions in parallel, emphasizing verification through behavior and documents rather than code review. The practitioner outlined rules …
An engineer detailed an architecture for running 10,000 AI agents without incurring LLM token costs, using the Agent Development Kit's 'before model' callback to intercept invocations and return deterministic tool calls.…
A developer's team at an unnamed company built a predictive autoscaling system for GPU workloads on Kubernetes after a production outage caused by reactive scaling. The system uses a Bi-LSTM model embedded in a Kubernete…
An engineer who ran an autonomous software team with 26 agent roles inside a GitHub repository for five months reports that the system merged 1,135 pull requests. The engineer found that AI code review by the same model …
In a technical blog post, the team behind xplainable, an ML platform, details how they built evaluation harnesses for their MCP server to test multi-step agentic workflows, arguing that early evals accelerate development…
Latchkey has introduced self-healing CI that automatically repairs environment-related build failures, while handing off code-related failures to AI coding agents via the Model Context Protocol. The system distinguishes …
A developer built a caching mechanism for Claude Code hooks to prevent performance degradation when checking token usage. The solution, cost_guard.sh, caches ccusage results for 60 seconds, throttling API calls to once p…
A developer detailed a production-grade MLOps pipeline architecture on AWS, using native services such as SageMaker, Step Functions, and CodeCommit to automate continuous training, governance, and canary deployments with…
A developer's guide to MLOps highlights the importance of deploying, monitoring, and optimizing machine learning models in production, emphasizing that models degrade over time due to data drift. The article covers GPU s…