Author: Danilo Poccia

Deploy AI agents with confidence using new quality evaluations and policy controls—enabling precise boundaries on agent actions, continuous quality monitoring, and experience-based learning while maintaining natural conversation flows.

New fast, cost-effective model supports extended thinking with adjustable reasoning depth, letting you control the balance between speed, intelligence, and cost while building AI applications for everyday workloads.

New program gives organizations unprecedented access to Nova model training, enabling them to build custom frontier models that deeply embed domain expertise without the traditional barriers of cost, compute, and time.

Build more natural voice interactions with enhanced speech-to-speech AI—now featuring multilingual conversations, dynamic speech control, crossmodal inputs, and improved telephony integration while maintaining conversation context across tasks.

Amazon Nova Multimodal Embeddings is a new state-of-the-art embedding model available in Amazon Bedrock. It is the industry’s first embedding model that supports text, documents, images, video, and audio through a single unified model, enabling cross-modal retrieval and semantic search applications with leading accuracy, at industry-leading costs.

This week I was at the inaugural AWS AI in Practice meetup from the AWS User Group UK. AI-assisted software development and agents were the focus of the evening! Next week I’ll be in Italy for Codemotion (Milan) and an AWS User Group meetup (Rome). My sessions there will be about AI agents and context […]

Amazon Bedrock has expanded its model offerings with the addition of Qwen 3 foundation models enabling users to access and deploy them in a fully managed, serverless environment. These models feature both mixture-of-experts (MoE) and dense architectures to support diverse use cases including advanced code generation, multi-tool business automation, and cost-optimized AI reasoning.

Build responsible AI applications with the first and only solution that delivers up to 99% verification accuracy using sound mathematical logic and formal verification techniques to minimize AI hallucinations and data ambiguity.