Perplexity's Search as Code Lets AI Models Build Custom Search Pipelines
Perplexity's new 'Search as Code' architecture allows AI models to write their own search workflows, reducing token usage by 85% in cybersecurity research tasks.
262 articles
The tools, frameworks, and platforms developers use to build with AI — coding assistants, agent frameworks, APIs, and developer tooling, with the updates that change how you ship.
Perplexity's new 'Search as Code' architecture allows AI models to write their own search workflows, reducing token usage by 85% in cybersecurity research tasks.
HuggingFace's new job search assistant uses Qwen3-8B and DeepSeek V4 Pro to generate tailored job recommendations based on resumes, with results delivered in a shortlist with reasoning.
Hugging Face's simulation shows a 3B model can create a functioning economy with five woodland agents, generating 100% valid JSON actions and market dynamics.
Microsoft's stock has fallen this year while competitors' valuations have risen, raising concerns about its position in the agentic AI race.
Mira Murati, CEO of Thinking Machines Lab, made her first major media appearance in 18 months during a Bloomberg interview, previewing new 'interaction models' designed to process continuous streams of audio, text, and video in 200-millisecond intervals.
Poke, a startup, has become the first AI agent approved to run on Apple’s Messages for Business platform, sending 100 million messages since its March launch.
Meta introduced an AI creator assistant on Facebook, now available to U.S., Canada, and India creators, offering personalized content suggestions and performance insights.
Microsoft announced Microsoft Discovery is now generally available, offering a platform for building and governing agentic AI workflows in scientific and engineering research.
Microsoft announced a new agent platform designed to integrate AI into enterprise workflows, enabling real-time, context-aware operations across multiple functions.
Amazon Bedrock now offers Ops Alert, a three-layer monitoring solution, to help organizations manage generative AI workloads at scale, reducing manual overhead and improving response times.
AWS announced support for SOCI snapshotter and index in DLAMI and DLC, reducing container startup times from 6 minutes 59 seconds to 21 seconds.
AMD released a new tool that allows users to fine-tune pre-trained LLMs for specific tasks without coding, using its AI Workbench platform.