engblogs

FinOps guide to updated spend-based CUDs: migration from credit-based to direct discounted pricing, validation via CUD Analysis, and enhanced cost reporting and scenario modeling for clearer savings.

Gemini 3 Deep Think: a technical exploration of advancing science, research and engineering.

Markdown for Agents enables token-efficient AI-ready content delivery by converting HTML to Markdown on the fly via content negotiation, empowering AI crawlers and agents to consume structured pages.

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.

Closing the evaluation gap in agentic AI is advanced through Open Benchmarks Grants funding for open benchmarks that quantify environment complexity, autonomy, and output complexity to enable safer, trustworthy AI agents.

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.