Paul Ogenekewen Emurotu

Led end-to-end AI/ML projects, fine-tuned domain-specific LLMs, built a scalable Python backend with LangChain and RAG pipelines, and implemented RLHF/SFT for code generation. Designed evaluation pipelines, deployed reproducible LLM CI/CD with Docker/Kubernetes, and built Gemini API wrappers for low-latency access under concurrent load. Led model evaluation and benchmarking with structured rankings for multi-response code outputs and delivered code quality assessments with detailed rationales.

Designed and implemented end-to-end data pipelines, data warehouses, and machine learning models to support client decision-making and operational efficiency. Built real-time dashboards and automated reporting systems to monitor KPIs. Developed advanced NLP solutions (Llama2, GPT-4, GPT-3.5-Turbo) and applied multilingual LangChain capabilities. Implemented semantic-search and recommendation systems, and deployed scalable ML solutions on AWS and Azure.

Delivered tailored ML solutions across NLP, image recognition, and graph mining, boosting model efficiency by 20%. Facilitated AI/ML training programs for 200+ learners. Built fraud detection systems achieving a 22% improvement in detection accuracy. Reduced client operational costs by 18% through strategic adoption of open-source ML and optimization of data workflows.