Oke Precious

I’m a Machine Learning Engineer who researches and builds agentic models, AI-powered workflows, and automation solutions. I focus on turning advanced ideas like Retrieval-Augmented Generation and LLM APIs into reliable systems that remove real engineering bottlenecks.

In my work, I design end-to-end pipelines and deployable services—from data modeling to production readiness. For example, I built a GDPR-compliant, privacy-preserving URL processing pipeline for 500,000+ URLs, eliminating web-scraping dependencies, and delivered an end-to-end Hybrid CNN-LSTM model for zero-day phishing detection.

I also optimize performance under concurrent load. In my ACE Bot, I architected a LangChain RAG pipeline using Flask and Docker to automate multi-modal document evaluation, reaching 99.9% pipeline uptime, and I engineered a real-time TTS audio service with sub-200ms streaming latency.

Beyond research, I care about measurable outcomes and transparency. My Credit-Scout Risk Engine uses an agentic LLM workflow with SHAP-based Explainable AI to generate interpretable reports for financial risk classification, and my hybrid e-commerce recommendation work delivered a strong Recall@10 with latency optimized for fast inference.