Been building a stock-sourcing tool for independent car dealerships (BMW/Audi/Mercedes-ish price range). Worth sharing since it overlaps with what's being discussed in the real estate/dealership thread.
What it does: scrapes Motorway, Carwow, AutoTrader and Facebook Marketplace against a dealer's buy-box (make/model/year/mileage/price), runs every match through the DVLA MOT History API (free, public) to flag mileage inconsistencies or heavy advisories before the dealer even looks at it, then fires a WhatsApp alert the moment something clears both checks.
The bit I'm most into: the dealer's own website has a "Notify Me" feature where customers register interest in a spec that's not currently in stock. That demand signal feeds back into the scraper, so instead of just finding cars that might sell, it prioritises cars that customers are already waiting for.
Stack: Apify (scraping), n8n (orchestration), Google Sheets (buy-box + CRM), Twilio (WhatsApp), DVLA MOT API.
Biggest challenge so far: matching listings against a buy-box reliably across sources with inconsistent formatting (mileage/price written differently site to site). Curious if anyone's solved that more elegantly than string-matching + tolerance ranges.