Miracle Chris-Mba
Machine Learning Engineer · Applied AI · Forward-Deployed Systems
I build applied ML systems that ship: computer vision, generative AI, and the backend plumbing that makes it all reliable. I like working on messy, customer-facing problems, and turning them into accessible, production-ready tools with real impact, all in service of engineering solutions for the laziest.
Things I've Done
ONA — Object Navigation Assistant (Assistive AI)
- Built an accessibility-first, real-time spatial awareness system for visually impaired users with a safety-first design mandate.
- Integrated YOLOv8, MiDaS, RetinaFace + FriendNet, and OCR into a multi-modal perception pipeline.
- Designed priority-based hazard detection and depth-aware distance categorization with low-latency (<500ms) guidance.
- Benchmarked custom-trained models vs transfer learning across accuracy, latency, and efficiency.
FriendNet — Multi-Model Image Classification Platform
- Built a FastAPI-based inference system serving multiple models (Custom CNN, ResNet, EfficientNet, CLIP).
- Fine-tuned OpenCLIP (ViT-B/32), first training heads then unfreezing full models for improved accuracy.
- Implemented evaluation pipelines including confusion matrices and robustness testing to guide iteration.
Where I've Been
AI Engineer (Co-op) — Ontario Public Service
- Reduced security incident triage time by 35% by designing and deploying a Bedrock-powered RAG system.
- Built LLM agents to crawl DataLake logs, update incident playbooks, and improve automated response time by 15%.
- Owned end-to-end delivery while translating ambiguous stakeholder needs into production AI systems.
Backend Engineer — SHIIP
- Improved logistics platform performance by 40% by redesigning rate and pricing services.
- Integrated third-party logistics APIs, increasing workflow reliability by 20% while handling edge cases.
- Implemented graceful degradation strategies, reducing geocoding-related errors by 15%.
Co-Founder & Backend Lead — Copnow
- Co-founded an e-commerce marketplace; acted as technical owner of backend systems.
- Led a cross-functional team of 3–5 engineers across product, infrastructure, and analytics.
- Integrated predictive analytics for order tracking, improving uptime by 30% and reducing errors.
Education
Artificial Intelligence (Co-op Program) — Fanshawe College
GPA: 4.2 / 4.2 · Computer Vision, Generative AI, ML deployment
BSc Computer Science — Madonna University
Thesis: Sensor-guided autonomous vehicle using Arduino and ultrasonic sensors
Writing
A personal tour through the words that ramble in my head, wrangled into stories about tech, leadership, and all the chaos in between.
Loading latest story…
View all stories