A hospital's AI vendor said the tool was "fully compliant" and "healthcare-ready". That was the pitch, spoken-out loud in the sales meeting. The actual contract said something else. Customer data could be used to improve the vendor's own model. Outputs were provided "as-is", with no warranty of accuracy or bias mitigation. Vendor liability was capped at one month of fees. None of that was hidden exactly, it just wasn't what got said in the room. I'm Boudhhayan Duttaa, founder

We started this one with a brand-new ad account. Zero history zero pixel data and zero excuse to hide behind the line "the algorithm just needs time to learn." That is the version of this story we actually want to tell because it is the hard one. Butterfly Learnings came to us as a healthtech buyer and healthtech buyers do not evaluate an agency on brand awareness or a nice-looking feed. They ask for a cost-per-lead number and a customer acquisition cost they can defend to th

Most clinic owners think they already fixed this. They added WhatsApp reminders eighteen months ago, no-shows dropped and the case closed, in their head. What almost nobody has checked since, is whether the fix itself is quietly creating a second problem, one that doesn't show-up on the appointment book at all. I'm Boudhhayan Duttaa, founder of Batti Jalao, an AI-led healthcare marketing agency. We build content and marketing systems including BattiLynk AI, BattiSense and cus