Bring superhuman perception to your product with computer vision built end-to-end — from research to deployed API
Production computer vision across medical imaging, agriculture, sports, and video analytics
Medical / AgriTech · USA
Multi-model system for bovine embryo grade, stage, pregnancy, and sex prediction from 30-second time-lapse clips — designed around each prediction goal with production-ready validation.
Instance segmentation of 11 artifact types in whole-slide pathology images, rebuilt with YOLO11x-seg to beat the client baseline
Single-recording padel analytics with RF-DETR player segmentation, ByteTrack IDs, homography, and a scaled 2D tactical map
Computer-vision pipeline for automated equestrian performance analysis, combining segmentation, motion recognition, and 3D vision
4-channel RGBN detection pipeline for aerial livestock monitoring: patch-embedding surgery, calibrated band normalization, and threshold-sweep evaluation dashboards
Real-time and post-event human movement analysis that classifies people as moving, stopped, or under-processing for security-focused video analytics
Non-invasive seed-group weighing from a single image using detection, segmentation, morphology features, and an XGBoost regressor
.gif){kind=avatar}
I'm Nassim — a computer-vision engineer from Tunis, hooked on everything a camera can measure that the human eye can't. Independent consultant and founder of CalibrX, I ship vision systems end-to-end — from messy raw footage to a deployed API. Grading embryos, weighing seeds from a single photo, counting cattle from the air, tracking four padel players through a glass box: if it needs superhuman perception, that's my kind of problem.