engblogs

summaries of the latest blog articles from your favorite tech companies.
Google CloudGoogle Cloud

Building Distributed AI Agents

Orchestrates a distributed team of specialized AI agents—researcher, judge, and orchestrator—into scalable microservices deployed on Cloud Run, enabling seamless frontend integration via the ADK and A2A protocol.

3/18/2026
AWS MLAWS ML

Introducing Nova Forge SDK, a seamless way to customize Nova models for enterprise AI

Nova Forge SDK unifies end-to-end LLM customization for enterprises, enabling developers to prepare data, train models, and deploy Nova variants across Amazon Bedrock and Amazon SageMaker AI with guided workflows and scalable automation.

3/18/2026
AWS MLAWS ML

Kick off Nova customization experiments using Nova Forge SDK

An end-to-end, Nova Forge SDK–driven guide to LLM customization—from data loading and supervised fine-tuning (SFT) to reinforcement fine-tuning (RFT), evaluation on a Stack Overflow classifier, and deployment on SageMaker AI Inference.

3/18/2026
Apple MLApple ML

Prose2Policy (P2P): A Practical LLM Pipeline for Translating Natural-Language Access Policies into Executable Rego

An end-to-end, LLM-driven pipeline that translates natural-language access policies into executable Rego (OPA policy-as-code), featuring policy detection, extraction, schema validation, linting, compilation and automated testing, demonstrated on the ACRE dataset with high compile and test pass rates to enable Zero Trust and auditable policy enforcement.

3/18/2026
AWS MLAWS ML

Migrate from Amazon Nova 1 to Amazon Nova 2 on Amazon Bedrock

End-to-end guidance for migrating from Amazon Nova 1 to Nova 2 Lite on Amazon Bedrock, detailing feature upgrades, API changes, migration steps, and production-ready evaluation to maximize reasoning, tool use, and throughput.

3/18/2026
AWS MLAWS ML

How Bark.com and AWS collaborated to build a scalable video generation solution

How Bark.com and AWS engineered a scalable, AI-powered video generation pipeline using SageMaker, Bedrock, and multi-GPU inference to automate creative ideation, ensure brand consistency, and accelerate content production.

3/18/2026
AWS MLAWS ML

Build an AI-Powered A/B testing engine using Amazon Bedrock

AI-powered, adaptive A/B testing on AWS using Amazon Bedrock and MCP to deliver real-time personalization with a serverless, scalable architecture.

3/18/2026
AWS MLAWS ML

Evaluating AI agents for production: A practical guide to Strands Evals

A practical guide to production-ready AI agent evaluation with Strands Evals, showing how to define cases and experiments, deploy LLM-based evaluators, and validate online and offline, multi-turn agent behavior.

3/18/2026
StripeStripe

Introducing the Machine Payments Protocol

Machine Payments Protocol (MPP) is an open, Stripe-enabled standard that lets autonomous agents transact programmatically, enabling microtransactions, recurring payments, and agentic commerce across fiat and crypto.

3/18/2026
Google CloudGoogle Cloud

The Proliferation of DarkSword: iOS Exploit Chain Adopted by Multiple Threat Actors

An in-depth examination of the DarkSword iOS exploit chain, its six-stage delivery and final payloads (GHOSTKNIFE, GHOSTSABER, and GHOSTBLADE) deployed across UNC6748, PARS Defense, and UNC6353 in watering-hole campaigns and loader-based delivery.

3/18/2026
MIT AIMIT AI

Sustaining diplomacy amid competition in US-China relations

Navigating competitive US-China relations to advance the energy transition and climate goals through diplomacy and strategic engagement.

3/18/2026
CloudflareCloudflare

Introducing Custom Regions for precision data control

Explores how Custom Regions extend Cloudflare's Regional Services to enable precise data localization with in-region TLS termination and L7 processing, while preserving global DDoS protection and data sovereignty.

3/18/2026