OpenAI Launches GPT-Live for Real-Time Voice AI
OpenAI unveiled GPT-Live, a real-time voice AI system, on August 3, 2026, enabling seamless, continuous conversation without turn-based interruptions.
Software
489 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.
LATEST NEWS
-
-
Amazon Bedrock Introduces Automated Reasoning Policy Refinement
Amazon Bedrock now offers automatic policy refinement, achieving up to 99% verification accuracy in automated reasoning checks, according to a recent announcement.
-
Formula 1 Uses Agentic AI on AWS to Cut Data Onboarding Time
Formula 1 reduced data source onboarding from 6 to 8 weeks to about 40 minutes using agentic AI on AWS, according to a new blog post.
-
OpenAI Presence Aims to Make AI Agents Production-Ready for Businesses
OpenAI's Presence offering targets enterprise customers, aiming to deploy AI agents in customer service and internal workflows by 2026.
-
AI Coding Agents Modernize Research Software but Can't Judge Scientific Accuracy
A field report from OpenAI and academic partners shows AI coding agents can speed up research software by over 60 times, but they can't verify scientific correctness.
-
Amazon Introduces Agentic Catalog Experience in Amazon Quick
Amazon Quick now enables seamless integration with upstream data catalogs, reducing time to insight from weeks to minutes, according to a recent blog post.
-
Amazon Bedrock AgentCore Observability Enhances AI Agent Performance
Amazon Bedrock AgentCore Observability helps identify and resolve performance bottlenecks in AI agents, with P95 response times exceeding thresholds leading to user abandonment.
-
Amazon Introduces Inference Meta-Monitoring for SageMaker AI Endpoints
Amazon announced a new system to track AI model performance in real time, with a 15% simulated drift in fraud detection scenarios.
-
Yahoo Uses Amazon Bedrock to Improve Search Retargeting
Yahoo enhanced its search retargeting capabilities using Amazon Bedrock, boosting keyword expansion rates up to 600 times in Q1 2025.
-
Intel DFlash Boosts Qwen3.6 Performance on AI PCs
Intel's DFlash method accelerates Qwen3.6-35B-A3B on Panther Lake laptops, achieving 2.2x speedup on HumanEval tests.
-
MCP protocol updates aim for enterprise scalability
The Model Context Protocol (MCP) received its largest update since its launch, introducing a stateless core to improve scalability for enterprise use.
-
Amazon Bedrock Introduces Advanced Prompt Optimization Tool
Amazon Bedrock now offers Advanced Prompt Optimization, a tool that optimizes prompts for up to 5 models while comparing original and optimized performance.