Bean On Bar
01 — THE PRODUCT
Bean On Bar turns scattered coffee information into one useful flow—before you buy, while you brew, and long after the bag is empty.
Photograph one or more bag angles and pull out the roaster, bean, origin, farm, variety, process, roast level and date, tasting notes, price, and weight. OCR runs on-device, then every field stays editable before scoring or saving.
Transparent signals from the label and evidence you paste—not a mysterious verdict or fabricated review.
Start from an unknown, identified, or saved bean. Choose Bean On Bar recommendations, your own recipe, or a roaster-style recipe; scale the dose, follow timed steps, and save the result back to the bean.
Save beans, ratings, recipes, price-per-gram history, tasting notes, watchlist matches, and journey recommendations. Back up and restore your notebook as a file when you need to move devices.
APP CAPABILITIES
Bean On Bar is still a prototype, but it already treats accuracy, transparency, and user control as part of the product—not afterthoughts.
Self-hosted Tesseract OCR reads each added photo in the browser, combines the text, and uses careful fallbacks instead of uploading bag images to a server.
The parser now handles coffee names, farm/region/country splits, roast dates, tasting-note labels, known OCR variety misreads, and multi-variety labels like SL28 + SL34.
Price per gram is compared against the user’s own saved history in the same currency, with ratings and “brew again” actions tied to real saved beans.
The home screen can surface the current bean and open the recipe timer immediately, so the app feels useful even before a large notebook exists.
Espresso gets its own recipe styles, bean-profile tuning, dose/yield controls, grinder guidance, and dial-in advice instead of being squeezed into filter recipes.
Travel mode routes users toward regional coffee sources, nearby cafe research, and “what’s good now” community examples while clearly marking static prototype data.
Export and restore local Bean On Bar data as a file, so the app can stay account-free while still giving users a way to protect their coffee memory.
Users can opt in to anonymous product events that improve the flow, while coffee notes, label text, photos, prices, and locations stay out of analytics.
CONNECTED COFFEE DATA
OCR starts the experience. The lasting value comes from connecting provenance, price, availability, roaster recipes, brew outcomes, backups, and trust signals into one useful bean record.
Explore Bean Intelligence