Mahmud Lawal.

Event-driven surveillance with a PIR + Raspberry Pi camera wake-trigger, YOLO object detection via the OpenCV DNN module, low-latency edge inference, and real-time email threat alerts.

Solved the “borderless office” problem with a dual-chatbot system for secure employee access: a WhatsApp-integrated assistant plus an identity-verification chatbot. Awarded overall winners.

Visual-first product discovery engine using vision-language models (Llama Vision) to extract features from product images, with OpenCV normalization and a LangGraph + FastAPI backend.

AI career coach built on FastAPI + Meta Llama 3: context-aware chat, a CV-tailoring engine that parses PDFs for gap analysis, ATS score estimation, and skill categorization, all containerized with Docker.

Spearheaded the control algorithm for autonomous navigation and obstacle avoidance. First team from Sub-Saharan Africa to compete, and secured a global leaderboard placement.

End-to-end video analytics pipeline classifying urban traffic congestion: YOLOv8 for real-time tracking, LightGBM for classification, with temporal lag/delta features for traffic-momentum accuracy.

YOLO-based computer-vision system for crop-disease identification: multi-stage segmentation and classification with a robust preprocessing pipeline, optimized for offline edge detection on the farm.