TaxiIntel

EQUINOX TaxiIntel is an AI-powered positioning intelligence system that increases taxi driver revenue by up to 31% per shift. I built an app for taxi drivers. Not because the market was obvious. Because the problem was beautiful. A taxi driver makes decisions every 4 minutes. Stay in this zone or reposition? Take the next short trip or wait for a long one? Is it worth driving 3km to the airport rank right now, or does the fuel cost eat the margin? Each decision has 6–8 competing parameters. Simultaneously. Fuel burn rate at current traffic density. Zone demand probability in the next 15 minutes. Tip likelihood by neighbourhood, time of day, and trip type. Start-fee leverage — the flat €5.00 that makes three short trips worth more than one medium one. Event proximity windows. Dead-mileage break-even. No human brain processes all of that in real time, against each other, correctly, every shift. The driver makes the call on instinct. Usually the wrong one — not because they're bad drivers, but because the data doesn't exist in a usable form at the moment of decision. That's the gap. Not a product gap. A cognitive one. What fascinated me was that every single parameter has weight — and the weight shifts depending on what the others are doing. High event day? Distance matters less. Fuel prices spike? Short-trip accumulation strategy flips. Midday lull? Zone matters more than trip type. It's a living system. The optimal decision at 11:47 is different from the optimal decision at 11:52. Building the algorithm that holds all of that — and collapses it into one instruction for the driver — turned out to be the most interesting engineering problem I've worked on. The result: +31% revenue per shift in simulation. Live in Düsseldorf.