Everyone else lists cars. We know what they'll actually do.
A brochure range figure is measured at 23°C in a lab. A listing price is whatever the seller typed. EV Cozmo replaces both with numbers derived from the region's own data — and every driver who uses the app makes them sharper.
One loop, four surfaces.
The same data that answers a driver's question also prices a seller's ad and tells an operator where to build.
What we collect
Marketplace index
1,200+ live listings, 5 sources, daily
Station telemetry
300+ stations, 9 CPO networks
Completed trips
Start SoC, arrival SoC, ambient temp
Our own range tests
8 cars measured at 45°C
Search & enquiry signals
What buyers look for, per region
What we compute
Range model
temp · speed · elevation · AC load
Price model
spec · mileage · region · days-to-sell
Demand model
searches vs. supply, by city
EV Assistant
answers grounded in the three above
Where it shows up
Route planner
arrival SoC & charging stops
Listing pages
real range + price guidance
Seller & agency tools
suggested price, demand heatmap
In-app assistant
"can I make it on 60%?"
CPO Yield & APIs
where to build, what to charge
↻ driver rates the trip plan · earns 10 points · model re-fits
The dashed return path is the only part competitors can't copy by scraping. Rating a route plan already earns 10 points in the app today — that reward exists to buy training data, and every rated trip tightens the range model for the next driver.
Real-world range
ShippingEvery range figure on the platform is adjusted for the conditions the car will actually be driven in. The route planner already takes an expected temperature and applies the penalty; the same model rewrites the number on every listing card.
Kia EV6 · brochure
528 km
At 42°C, 120 km/h, AC on
431 km
Price intelligence
NewWhen a seller reaches the price field, we already know what the same spec sold for, how long it took, and what's listed right now. Suggested price appears inline — and the seller sees the trade-off rather than a single number.
Suggested for your EV6 GT-Line
2.05M · sells in ~4 days
2.25M · ~12 days
2.6M · 60+ days
Demand intelligence
NewWe see the searches that return nothing. That gap — demand with no supply — tells an agency what to import next and a charge point operator where a station would pay for itself.
Unmet demand · Cairo, last 30 days
EVs under 1M EGP
2,410
7-seat electric
1,180
Electric pickups
604
EV Assistant
ShippingA general chatbot guesses. Ours answers from your car, your battery, today's weather and the live station map — so the answer is specific enough to act on, and it can hand off straight into a planned route.
Can I make it to Alexandria on 60%?
Yes — 303 km usable at today's 38°C against 220 km of route. You'd arrive on about 22%.
What exists today, and what doesn't
Route planner with expected-temperature input and a battery model
Trip feedback loop rewarding 10 points per rated plan
features/ai_agent — the assistant chat surface
Marketplace index across five sources, with a filter taxonomy
The price model and suggested-price slider
Unmet-demand reporting for agencies and CPOs
Heat-adjusted range shown on listing cards, not just in routing
The 45°C range-test dataset itself — the tests have to be run