Mercy Ogubuike

Created and completed real-world Adobe Illustrator design tasks for OpenAI model training and evaluation. Recorded end-to-end workflows and produced multimodal training data.

Led 0→1 product engineering connecting model alignment with production full-stack architecture. Built multi-turn AI systems using Python and React.js with persistent context and real-time model interactions.

AI training engagement focused on model-oriented task execution and evaluation.

Full-stack AI engineering engagement supporting AI-focused product and laboratory workflows.

Managed technical recruitment and vetting for 120+ engineering hires across backend, DevOps, AI, LLMOps, and cloud roles. Conducted technical screens and built sourcing pipelines.

Led production AI delivery across automotive, healthcare, fintech, and e-commerce; owned training/serving stacks, CI/CD, model governance, and multi-cloud deployments. Built standardized fine-tuning pipelines and operationalized vLLM/Triton serving on Kubernetes.

Owned the perception and sensor-fusion roadmap for ADAS with a sub-50 ms on-vehicle latency target; led a 5-engineer lab. Delivered camera + LiDAR fusion using EKF/UKF and learned fusion, reducing testing incidents by about 50%.

Resolved Tier 2/3 technical escalations involving software bugs, API authentication errors, and database inconsistencies. Automated repetitive support workflows using Python scripts and Zapier, reducing manual ticket processing time by about 80%.

Trained and deployed NLP and computer-vision models for classification, sentiment, and image tasks, achieving up to 40% production accuracy improvement. Implemented SFT/RLHF pipelines and evaluation harnesses.

Led TDD-first backend delivery with 90%+ test coverage, CI gates, and contract tests. Implemented Prometheus/Grafana monitoring, reducing downtime by about 20% and MTTD/MTTR by about 15%.

Built a lightweight EMR backend with Django/Flask and REST APIs for patient registration, appointments, labs, and billing as a community-service project. Automated lab-result ingestion and validation, reducing duplicate entry and data errors by about 70%.

Completed the Machine Learning with Python course covering linear models to deep learning.

Earned a Diploma in Python for Data Science.

Completed an associate-level course in Machine Learning and AI Libraries.

Completed the Minds and Machines course.

Completed an associate-level course in LLM and Deep Learning.

Grade: Second Class Upper

Earned a Bachelor of Science in Architecture with Second Class Upper honors.