Chidera Obiezue

I am an ML and backend engineer who builds production-grade MLOps pipelines, scalable APIs, and data-driven web systems. I blend practical engineering with applied ML to deliver reliable, cost-efficient solutions.

I've engineered multi-stage LLM pipelines, designed asynchronous task queues, and implemented cursor-paginated DRF APIs that eliminate N+1 query problems at scale. My projects include a scholarship discovery platform and a platform-agnostic moderation system deployed to serverless infrastructure.

In a moderation project I combined DistilBERT with XGBoost to cut false positives and achieved a 92.4% F2 score, implemented multi-stage drift detection, and reduced inference costs through caching and optimized feature stores. I also deployed containerized models to AWS Fargate with automated CI/CD and monitoring.

As a software intern, I built data engineering pipelines for image and video annotation, authored annotation guidelines that improved labeling consistency, and performed QA on FastAPI ML endpoints. I focus on measurable impact, robustness, and maintainable systems.