Skip to content
View ayodeji07's full-sized avatar

Block or report ayodeji07

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ayodeji07/README.md

Hi, I'm Ayodeji Akande 👋

HealthTech Data Scientist & AI Engineer · Lagos, Nigeria

I'm a human anatomist turned data scientist — which means I don't just work with health data, I understand what it represents biologically and clinically. I'm currently completing an M.Sc. in Public Health, and I build end-to-end data and AI solutions at the intersection of medicine and technology.

🎻 Outside of data, I'm a violinist. The precision and pattern recognition that classical music demands is exactly what I bring to every dataset.


🔭 What I work on

  • NLP & LLMs — fine-tuning biomedical language models, clinical NER, ICD-10 mapping, RAG pipelines
  • Computer Vision — deep learning for medical imaging and image-based diagnostics
  • Health data engineering — ETL pipelines, surveillance systems, geospatial analysis
  • Clinical research analytics — statistical modelling supporting 40+ peer-reviewed research projects

🛠️ Tech Stack

Languages & Analysis

Python R SQL Power BI

Machine Learning & AI

scikit-learn XGBoost PyTorch LLMs & RAG

Data Engineering & APIs

FastAPI PostgreSQL Docker GitHub Actions

Dashboards & Visualisation

Streamlit Plotly


🏥 Projects

Computer Vision Deep Learning Medical Imaging

Multi-label classification of 10 chest pathologies using DenseNet121 (CheXNet architecture) trained on NIH ChestX-ray14. Includes Grad-CAM heatmaps showing which image regions drive each prediction, and a live Streamlit demo.

Demo Model


NLP NER BERT Clinical Text Mining

Production-grade pipeline that extracts structured clinical knowledge from unstructured medical notes — diagnoses, medications, procedures, symptoms, and anatomical terms — using scispaCy and fine-tuned biomedical BERT, with automatic ICD-10-CM code mapping.

Demo


Survival Analysis Machine Learning R SHAP

Full clinical research pipeline combining R survival analysis (Kaplan-Meier, Cox PH) and Python ML (XGBoost, LightGBM) for patient mortality prediction on the Worcester Heart Attack Study (WHAS500), with SHAP explainability at both global and per-patient level.

Report


Data Engineering FastAPI PostGIS Streamlit

End-to-end platform tracking 5 diseases across 37 states (2015–present). ETL pipeline parses 155+ NCDC PDFs into a PostGIS database, a FastAPI serves 20+ analytical endpoints (trend tests, outbreak detection, forecasting), and a live dashboard renders geospatial and statistical insights.

Dashboard API


🌱 Currently building

  • Building expertise at the intersection of human anatomy, computer vision, and AI-assisted medical imaging
  • Exploring robotics and multimodal models that combine anatomical knowledge with visual intelligence
  • M.Sc. Public Health dissertation (LAUTECH, 2026)

🤝 Let's connect

Open to remote and global opportunities in HealthTech · Health Data Analytics · ML/AI Engineering · Clinical Research.

LinkedIn Email

Pinned Loading

  1. chest-xray-classifier chest-xray-classifier Public

    Multi-label chest X-ray pathology classifier (DenseNet121) with Grad-CAM explainability, on NIH ChestX-ray14

    Jupyter Notebook

  2. clinical-nlp-pipeline clinical-nlp-pipeline Public

    Clinical NLP pipeline: scispaCy NER, ICD-10 mapping, and Bio_ClinicalBERT severity classification

    Jupyter Notebook

  3. nigeria-disease-surveillance nigeria-disease-surveillance Public

    Disease surveillance platform for Nigeria: ETL, spatial analysis, outbreak detection, and forecasting

    Jupyter Notebook

  4. patient_survival_prediction patient_survival_prediction Public

    Survival analysis (Cox, Kaplan-Meier) and ML on the Worcester Heart Attack Study, with SHAP explanations

    Jupyter Notebook