Barista feedback system that spots inconsistencies before customers do

Cafés bleed loyal customers without ever knowing it. Nobody explains their departure — they just quietly disappear because the cappuccino they loved on Monday was unrecognizable by Tuesday. One barista extracts shots in 25 seconds while another takes 35. Milk temperatures fluctuate by 20 degrees depending on who's working. Regulars pick up on this. They drift away for good. Pour Score is a computer vision platform that monitors every beverage as it's crafted, delivering real-time coaching to baristas before the drink ever reaches the customer. A compact camera installed above the espresso station detects over-extracted shots, poorly steamed milk, and sloppy pours as they happen. Think of it as a veteran barista hovering behind every trainee, quietly guiding them back to the standards that earned the shop its devoted following. The platform absorbs each café's unique recipe profiles and becomes the most efficient tool for onboarding new staff while keeping seasoned baristas dialed in and uniform. The core monitoring service starts at $29 per location per month. An analytics dashboard surfaces quality patterns, shift-by-shift consistency ratings, and flagged beverages that need work. Specialized training modules — covering latte art, pour-over methods, or seasonal menu launches — range from $200 to $500 each. The ideal entry point is independent cafés and small chains (3 to 10 locations) that are fanatical about quality yet can't justify a dedicated quality manager at every bar. These are owners who personally sample drinks throughout the day. Owners who dread Yelp reviews calling their coffee "hit or miss." Owners who recognize their regulars by face and order. The MVP keeps things lean. Install a $200 USB camera above a single espresso station in a beta location. Route the video feed to a laptop running AI Vision with custom prompts calibrated to evaluate timing, color, and technique. Build a lightweight web dashboard delivering live guidance: "shot pulled too fast," "milk temp looks low," "good pour, nice work." The objective is demonstrating that the AI reliably catches errors and that baristas actually welcome the input without feeling surveilled. Pilot with 3 to 5 friendly café owners who genuinely care about drink quality. A single location humming along for 30 days with quantifiable consistency gains is both proof of concept and a story that sells itself.

What you get with the full dossier