AI Infrastructure news and analysis on Web Pulse: 32681 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
A new empirical study found that layering three retrieval methods—typed discovery, semantic context, and file verification—achieved a 0.954 score for LLM-generated engineering artifacts, outperforming grep alone (0.918) …
Canonical and Arm have partnered to certify Ubuntu 26.04 LTS on the new Arm AGI CPU, a high-performance processor designed for agentic AI workloads. The Arm AGI CPU features up to 136 Neoverse V3 cores on a 3nm process w…
Agent Activity is an open-source monitoring platform that combines a Python endpoint agent, FastAPI backend, and Angular dashboard to collect and display system activity from registered desktop machines. The tool can cap…
A team building a self-improving knowledge base called Morsel developed a method to structure technical documentation for use by local AI coding agents. The process classifies pages using rules and a local LLM, embeds co…
The llm-cli-gateway v1.6.0 now implements first-class prompt caching across all five supported LLM providers—Claude, Codex, Gemini, Grok, and Mistral Vibe—allowing developers to avoid paying for repeated input tokens on …
CoreWeave CTO and co-founder Peter Salanki discussed the challenges of running AI in production at the HumanX conference, emphasizing the growing importance of observability, utilization, and scheduling. Salanki also adv…
Lore, a decision memory layer for projects built with Claude Code, captures the reasoning behind code changes automatically and locally alongside a project's git workflow. The tool records agent activity and decisions du…
A developer has introduced the AI Development Life Cycle (AIDLC), a structured framework designed to address the unique challenges of machine learning systems that traditional software development life cycles (SDLC) cann…
A new AI agent token cost calculator estimates monthly spending and waste for coding loops like Codex and Claude Code, allowing developers to input token counts, run frequency, and provider rates. The tool includes a was…
At Warmtebouw, a mid-sized engineering firm, a developer deployed an enterprise AI system without the typical costs of a custom chat UI, per-token API billing, or a dedicated platform team. By using a vendor's chat clien…
The VEKTOR project has released a 10-minute tutorial demonstrating its hybrid memory system, which combines local SQLite storage, AES-256 encryption, and semantic recall to transform AI assistants into stateful infrastru…
Warmtebouw deployed nine MCP servers in three months for non-technical business users across ERP, BIM, fleet, energy, and operational systems, bypassing the entire conventional AI stack of orchestration frameworks, vecto…
Cognicore has released NEXUS, an open-source framework that provides a standardized memory, retry logic, and orchestration layer for AI agents, eliminating the need for developers to rebuild these components from scratch…
Google engineer Eric Biggers posted proof-of-concept Linux kernel patches Monday adding support for ML-KEM and X-Wing, two post-quantum cryptographic standards designed to secure systems against future quantum computing …
Supabase reported a typosquat package named `supabase-javascript` on npm that mimicked its official package to phish developers, which npm removed hours later after it had already accumulated real downloads. The company …
A new open-source ecosystem called Riot launched for the OCaml programming language, providing a complete stack with a multi-core actor-model runtime, package registry, and standard library. The project ships with an age…
FolioDrop launched a deployment platform that lets users publish AI-generated content directly to the web as plain HTML pages via a single connection token. The service enables ChatGPT and other AI agents to return real …
Seven open-weight frontier models, including Hugging Face's SmolLM3, DeepSeek-R1, and OpenAI's gpt-oss-120b, were analyzed to distill common training methodologies for multi-billion parameter models, with a focus on arch…
Nexus Labs reduced time-to-first-token (TTFT) from 480ms to 110ms on one tenant by enabling vLLM's prefix cache for agent workloads, while another tenant saw no improvement. The discrepancy was caused by Tenant B's agent…
Dory has added support for Oracle Database, enabling users to connect and work with Oracle alongside other data sources in a unified workspace. The integration supports the full data workflow, including browsing schemas,…