ALAFIAI

The problem Point-of-care tests — glucose, hemoglobin, INR, HbA1c — are run where labs can’t reach: mobile clinics, rural health posts and community programs across low- and middle-income countries. Most of those devices are never quality-controlled. Strip lots drift and calibration lapses, and nobody is positioned to notice. A wrong INR at a rural health post can mean an unsafe anticoagulant dose, and often no laboratory scientist is nearby to catch it before the patient takes it. What POCT Hub does POCT Hub links mobile clinics to a central laboratory for remote quality control. Registered clinics send results and device photos from their POCT devices, including over a WhatsApp-style channel, and Meta Muse Spark reads the photos and printouts. Results then pass seven safety gates and Westgard QC rules. The hub tracks strip-lot drift and flags calibration problems, and anything critical goes to a human reviewer. If the AI is unsure or unavailable, it fails closed rather than guessing. Clinics get technical recommendations; the central lab gets one dashboard covering every device in the field. I’m a hematologist, and I built this for the settings I’ve worked in. It runs on synthetic data and is for research and decision support only.