Salman Guliwala
Technical leadership & business strategy
I lead the people who build it, and I explain it to the people who fund it.
Most technical work fails in the explaining, not the building.
I work where engineering, product, and the business meet: leading the build, owning the business case, and keeping all three talking about the same thing.
Demand forecasting, retail
Replaced a spreadsheet process nobody trusted with a model the planners actually open. The hard part wasn't the forecast — it was getting three directors to agree on what "accurate" meant.
Claims automation, insurance
Built the pipeline, then built the business case that got it funded — a one-page model of cost per claim before and after, which turned a nine-month debate into a decision.
Alfred
A personal agent that runs my mornings, holds my open loops, and argues with me when I'm about to make a bad decision. Where I test ideas before I bring them to anyone's budget.
AI for Everyone
Short-form video and writing that explains AI to people who don't work in it — the same translation problem as the boardroom, with a harder audience and no slide deck.
The ROI Conversation Engineers Avoid
Finance isn't asking you to dumb it down. They're asking which number moves, by how much, and by when — and "it depends" is not an answer, it's a refusal.
What Your AI Pilot Actually Costs
The line items nobody puts in the deck: evaluation, drift, and the two engineers you will quietly lose to maintaining prompts.
The Literacy Gap
The people most affected by this technology are the least likely to be in the room where it's explained. That's a choice, not a side effect.
What the Camera Taught Me About Attention
A frame is a decision about what to leave out. So is a roadmap, a board update, and most of a life.
If you're weighing up a piece of technology and can't get a straight answer on what it's worth, that's the conversation I like most. Email is fastest.