Omobuwa Blessed Adeoye
Available for AI engineering roles

Omobuwa Blessed Adeoye

Health Information Manager · AI Engineer · Data Engineer · Data Scientist · Ibadan, Nigeria

From health information management to AI engineering — I build agents that make clinical data useful.

About

I work at the intersection of health information management, data engineering,and applied AI. My day job is at University College Hospital, Ibadan, where I manage health data and build analytics that clinicians actually use.My nights and weekends go into agentic systems — language models that reason,use tools, and ground their answers in real sources. medagent is where those two threads meet: a healthcare agent that queries databases, generates charts,and cites WHO guidelines, all in one reasoning loop.

Based in
Ibadan, Nigeria
Focus
LLM agents · RAG
Stack
Python · Django · Plotly
Status
Open to opportunities
What I work with
AI & ML
LLM agents RAG Tool calling Prompt engineering scikit-learn Model evaluation
Data
Python pandas NumPy SQL Plotly Statistical analysis
Engineering
Django FastAPI Streamlit Docker basics Git REST APIs
LLM Stack
Ollama Cloud Groq ChromaDB sentence-transformers Llama GPT-OSS
Selected projects
medagent
Deployed

A Django + LLM agent that queries a healthcare database, generates charts, and cites medical guidelines. Built from scratch with native tool calling and RAG.

Django Ollama Cloud ChromaDB Plotly
Open project
COREP Outreach Analytics
Deployed

An end-to-end healthcare analytics platform for medical outreach programs — 46+ interactive charts, ML predictions, and an Ollama-powered AI assistant.

Streamlit scikit-learn Ollama Plotly
Open project
Medical Prediction Models
Research

Breast cancer and Parkinson's disease prediction using classical ML — feature engineering, model comparison, and evaluation.

scikit-learn pandas NumPy
No public link yet
Currently
Deploying medagent to Render with a pre-computed RAG index with evaluation, monitoring and free-tier-safe loading. Next up: streaming responses and fine-tuning a small model on the COREP medical domain.
Get in touch