/ Writing
Getting AI agents to do meaningful work inside a real company is mostly a data problem. Here's why 15 years of pipelines, models and governance turned out to be the perfect training ground.
For decades we argued about how BI tools should connect to data. AI agents bring the same question back, with higher stakes. The Model Context Protocol is the most promising answer I've seen.
Most AI initiatives stall for reasons that have nothing to do with the model. A practical look at the five data foundations that decide whether enterprise AI actually works.
Fifteen rounds of shouting "Cat!" at pictures, one tiger that split the room, and the lesson I most want people to take away about working with AI: compute is cheap, purpose is rare.
Ask an agent for 'revenue last quarter' and it has to guess what you mean. A semantic layer turns that guess into a governed answer, and it may be the most underrated piece of the AI stack.
The fastest way to lose trust in enterprise AI is one bad answer or one unauthorized action. A practical checklist for connecting agents to real data without creating a new risk surface.
Why Tableau + Snowflake is a match made in data heaven: columnar micro-partitions, result caching and elastic scaling that make live connections the right default.
50-page custom SQL scripts, Excel on SharePoint, cubes and more. Five habits that signal it's time to invest in a trusted, governed data platform.
Billion-dollar companies still move data with 40-year-old tech. A hands-on guide to sharing live, read-only data with outside organizations in under 15 minutes.
Married, moved across the world and started a new job in ten days. Then I dug into SingStat data to see how Singapore's real income distribution compares to the movie.