All work
PHP / Electron / SQLite

Replaced a paper referral log with a live map and a Telegram ping.

A Pretoria specialist practice was tracking referrals on paper plus a spreadsheet. No map of where referrers were located. No quick way to see which practices were sending the most patients. The front desk ran a separate prospect list that lived in another file and got out of sync.

Aerial view of a city at night with glowing street grids, abstract geographic pattern

Built a PHP and SQLite web app with a Leaflet map plotting referrer practices. Access is device-key gated: a shared link plus a token, no full login system to manage. When a new referral comes in, a Telegram bot pings the principal directly. POPIA shaped the data model from the start. Referral data is stored locally, access is scoped, no third-party analytics touching patient records.

For front-desk prospect management, a separate Electron desktop app handles the workflow that never belonged in the web app. It ingests prospects from four sources: a medical scheme directory, the regulator's registry, OpenStreetMap, and existing spreadsheet data via merge-import. Smart Paste pulls prospect details from copied text without manual re-keying. The local database is PIN-locked.

Principal has real-time visibility into the referral pipeline without opening a spreadsheet. Front desk manages prospects in one place. Replaced a paper-plus-spreadsheet workflow across both sides of the practice.

Stack: PHP, SQLite, Leaflet, Telegram Bot API, Electron, Node.js.

Back to all work   Start a similar project