The Premature Evals Trap: Why AI Metrics Are Lying to You
I watched a data team present their “AI evaluation framework” last month. Ninety percent accuracy on their test set. They were genuinely proud. Three weeks l...
Recent posts
The LLM Token Subsidy Whiplash: When AI Economics Meet Reality
The story of subsidized AI tokens is shaping up to be one of the most consequential economic whiplashes of the 2020s. We’re living through a pivotal moment w...
How GenAI is Reshaping Open Source Collaboration (And Why It Matters)
A few months ago, I was deep in a GitHub issue thread—one of those “I want to contribute but I’m completely lost” moments. The project was active, the docume...
Next 5 years – Another Raul’d-Map (2026 Roadmap)
Just like Donald Trump, here we go again…
Transforming Analytics Support With AI
As a data professional with deep experience, I lead a team for a short stint and built two analytics platforms in the era of “Modern Data Analytics”, focusin...