I spent a decade in rooms where capital allocation decisions got made — London deal teams at Platinum Equity and Terra Firma, project work at McKinsey on integration, cost, and commercial diligence, operational transformation at Yara, and finally investment management at Kistefos.
The drudge was always the same. Senior people spent too much time reconstructing facts from documents that juniors had already worked through. Review bottlenecks moved slower than deal volume. Most of what we called "analysis" was really retrieval, done twice.
I've also always had an entrepreneurial itch. I shipped side projects in the B2C productivity space while the day job paid the bills — React, Node, small apps, the usual — mostly to scratch it, sometimes to prove a point to myself.
What changed is that AI finally got reliable enough to take on the retrieval layer without making things up. Not "AI transformation" reliable — source-backed, traceable, reviewable reliable. That's the gap Verthandi AI exists to close: narrow tools for investors and advisors, built by someone who's done the work they're replacing.
The broader reason I care: capital allocation affects which companies get to grow, which don't, and what the economy looks like a decade from now. If AI can remove enough drudge that investors focus on the parts of the job that actually require judgment, more capital ends up where it's genuinely productive. A small social note, but it's why I'm building what I'm building, and not something else.