Insights
Notes from the work.
Practical writing on building enterprise AI that reaches production.
A fraud model is only as current as its features. Snowflake's Online Feature Store is now generally available
Snowflake made its Online Feature Store GA on AWS and Azure, serving the latest feature values to real-time models from batch, stream and real-time views. Freshness becomes something to monitor.
Read
Your HR agent gave the correct answer. Did it behave as your company requires? Cornerstone tests that
Cornerstone launched Agent Ready to evaluate how HR and customer-facing agents exercise judgment, not just whether they are accurate. Conduct is the evaluation gap most agent programs have not closed.
Read
Google is teaching Gemini government procedure. The design question is where a draft becomes a decision
Gemini for government services, in preview, pairs expert-written skills with agency connectors and partner agents. Each workflow needs a clear line between a proposal and a decision.
Read
Before an AI agent gets a role in your warehouse, can you say who that role already reaches? Snowflake now can
Snowflake made two Trust Center reports GA: who is entitled to tables with sensitive data, and who actually queried them. For agent access reviews, the gap between the two is the finding.
Read
Salesforce is paying $5.66 billion to act on customer behavior in real time, not for another AI model
Salesforce agreed to buy Braze, whose event-streaming platform turns live app and web behavior into messages across channels. For agentic marketing, the gap was never ideas. It was execution.
Read
Sovereign AI is not where your data sits. It is what keeps working when the connection is cut
Google now runs Gemini in air-gapped data centers and Microsoft is taking Foundry and GitHub Copilot offline. Both show sovereignty has become an operating question. Seven questions to test any offer.
Read
A voice agent is a real-time system, not a chatbot with a microphone. Mistral and monday.com build it two ways
Mistral's Voxtral 3 ships speech models and a modular voice agent SDK; monday.com's Harmony runs small specialized models behind a sales platform. Latency, not intelligence, shapes both.
Read
Gemini in Google Sheets now queries your whole BigQuery dataset. The spreadsheet became an analytics agent
Gemini now analyzes BigQuery data from Connected Sheets in plain language and builds refreshable dashboards. It is on by default, with no admin control. Definitions, cost and ownership need deciding.
Read
OpenAI bought Orchestra Bio to go from helping scientists discover drugs to coordinating their development
OpenAI acquired Orchestra Bio, a platform where people and agents track drug programs, to join Rosalind. Coordination agents in regulated programs raise questions research agents never did.
Read
Anthropic's automated red team beat the best human bug bounty finding in 23 of 24 tests. What does that prove?
Anthropic says an automated red-teaming system matched or beat human bounty hunters on every one of 24 classifier tests, in 24 to 48 hours. The result is real and narrow. Here is how to read it.
Read
The next cut in AI inference cost may come from below the model. llm-d 1.0 shows where
llm-d, the CNCF project backed by IBM, Red Hat, Google Cloud, NVIDIA and CoreWeave, reached 1.0 with tenant priorities, session-aware routing and zero-downtime upgrades, the cost drivers for agents.
Read
AI agent observability tools: what each one actually covers, and what none of them can tell you
Snowflake, AWS, Microsoft, Collibra and others now sell observability for AI agents. They cover different layers. A guide to the four layers, the tools in each and the questions to ask before buying.
Read
Have an AI use case stuck between prototype and production?
Tell us what you’re trying to ship. We’ll reply with honest next steps.
Discuss a use case