engblogs

Step-by-step guidance on correctly orienting a chessboard and placing all pieces—from pawns to the king—with beginner-friendly timing tips to enable repeatable, practical play.

GTIG's AI Threat Tracker analyzes how threat actors deploy AI—from distillation/model extraction attacks to AI-augmented operations and agentic AI—to accelerate reconnaissance, phishing, and malware development, while detailing Google's defenses and safeguards.

AI and physics-based simulations are accelerating materials discovery and scientific progress through high-throughput computation, generative modeling, and multimodal reasoning across MIT and beyond.

Databricks delivers gov-ready AI and data intelligence on AWS GovCloud with AI/BI Genie, the Foundational Model API, and Databricks Assistant, all governed by Unity Catalog and backed by serverless compute.

Delta Sharing enables secure, real-time data collaboration between retailers and CPG brands, replacing complex ETLs with instant access to sales, inventory, and product data to optimize promotions and prevent stockouts.

A pragmatic look at how domain intelligence defines what constitutes 'high quality' in production AI, with practical metrics, monitoring, and governance to deliver reliable, business-aligned outcomes.

February 2026 roundup showcases agentic analytics and end-to-end AI/BI workflows powered by natural-language dashboards across Databricks One, Genie, and Dashboards, with stronger governance via Unity Catalog and easier sharing via Teams and APIs.

A race-against-the-clock case study showing how Data 360 built an AI-ready, scalable knowledge base and launched the Informatica Help Agent in 24 days by unifying 100,000 unstructured documents into retrievable context through reusable patterns and disciplined ingestion and retrieval.

Details the $3M Open Benchmarks Grants to close the evaluation gap in agentic AI by advancing open benchmarks for safe, trusted autonomy in complex, real-world environments.

Advent of FPGA Challenge brought 213 submissions from 46 countries, showcasing diverse HDL approaches to solving Advent of Code puzzles with real hardware demonstrations.

AI-powered, context-aware tabletop exercises on Google Cloud help financial institutions validate regulatory readiness and strengthen operational resilience under DORA.

NVIDIA Nemotron 3 Nano 30B MoE model with 3B active parameters and a hybrid Transformer-Mamba architecture is now generally available in Amazon SageMaker JumpStart, enabling open-weights and fully managed deployment to accelerate AWS-based AI development.