Berkeley AI Research Lab (BAIR) celebrates 2026 Ph.D. graduates' impactful work in AI, robotics, language models, and more. Graduates head to academia, industry, and startups, shaping the future of AI.
Amazon Bedrock AgentCore Memory is a managed memory service that improves AI agents' ability to recall information accurately. Metadata filtering enhances retrieval precision, boosting overall question-answering accuracy from 40% to 64%, with significant gains in contextual questions.
Automated AI agents are on the rise, with 35% of businesses already using them and 44% planning to. MIT expert Phillip Isola explains agentic AI's actions in the world, differentiating it from generative AI like ChatGPT and Claude.
Manage AI model access across multiple AWS accounts using managed entitlements for Amazon Bedrock, eliminating the need for AWS Marketplace permissions in member accounts. This approach streamlines distribution, ensuring teams can access third-party models like Anthropic Claude or Cohere while maintaining centralized governance and avoiding operational overhead.
AI agents go beyond chat with AG-UI protocol, enabling interactive chart rendering and real-time updates. Amazon Bedrock AgentCore offers a secure platform for building AI agents with AG-UI integration, connecting backend frameworks with frontend libraries seamlessly.
AWS focuses on security with AI services like Amazon Bedrock, offering top performance and privacy. Anthropic's Claude Mythos models bring new cybersecurity capabilities, emphasizing the importance of balancing innovation with safeguarding against misuse.
Outpost VFX achieves 8x faster AI model training using AWS infrastructure, overcoming single-GPU limitations. Collaboration with AWS Generative AI Innovation Center modernizes face replacement workflow for high-end VFX production.
Singh leads NVIDIA Isaac ROS, enhancing robotics with CUDA libraries. Isaac ROS offers modular packages for perception, mapping, and motion planning, empowering developers in the evolving field of robotics.
Implementing L2 regularization in linear regression using MP pseudo-inverse via QR-Householder, with a demo showing modest improvements. The technique, though complex and rarely beneficial, can be applied to various linear models for experimental purposes.
Implementing resilience patterns for Large Language Model (LLM) inference is crucial as generative AI workloads scale. Amazon Bedrock offers features like cross-Region inference for maintaining availability, response time, cost, and throughput in production.
Anthropic releases Claude Sonnet 5, a highly agentic model surpassing Sonnet 4.6. Offers superior benchmarks at lower costs.
Fine-tune Amazon Nova models with Amazon SageMaker AI to automate accurate data extraction, reduce costs, and improve efficiency. Parcel Perform collaborated with AWS GenAIIC to optimize Nova models, achieving up to 94.77% extraction accuracy, reduced costs by 50%, and improved latency for ecommerce logistics operations.
Amazon Bedrock AgentCore Observability helps debug AI agent failures by providing visibility into agent execution, allowing you to trace reasoning steps and identify failures. Understanding quality, reliability, and efficiency failure patterns helps narrow down investigations quickly.
PAR Technology Corporation supports 300+ restaurants with data-driven decisions. Their innovative LLM analytics system enforces strict row-level security, ensuring accurate results for each user.
Efficiently digitize scanned yearbook pages with Amazon Nova 2 Lite and Anthropic’s Claude Sonnet 4.6, producing 3,122 accurate name-to-face associations at a lower cost than single-model alternatives. The two-model pipeline uses native multimodal extraction and spatial reasoning to match names to faces based on page layout, offering a cost-effective solution for large-scale document digitization.