Project Showcase

Exploring the intersection of innovation and technology

MindMate AI: An Agentic AI for Mental Health Triage

πŸ—‚οΈ AI Tools Completed
AI Agentic AI Mental Health Healthcare IBM Watsonx Python Flask JavaScript HTML CSS Triage Clinical Dashboard Accessibility Multi-language

πŸ“ Overview

MindMate AI is an intelligent assistant designed to support mental health triage in under-resourced community clinics. It features a user-friendly web interface for patient intake and a secure clinical dashboard for staff. The core of the project is an "agentic AI" powered by IBM watsonx.ai, which analyzes user-submitted symptoms to provide not only an empathetic recommendation but also a crucial severity assessment (Low, Medium, High). This allows clinic staff to instantly identify and prioritize high-risk patients, solving the critical problem of delayed care. The application is built with a robust Python/Flask backend and a polished, accessible frontend.

🌟 Impact

The project's primary impact is on improving the efficiency and effectiveness of mental healthcare in underserved communities. By automating the initial triage process and providing an intelligent severity score, MindMate AI helps reduce staff burnout, shortens wait times for critical patients, and ensures that the most vulnerable individuals receive timely care. Its multi-language and voice accessibility features further broaden its impact, making mental health support more inclusive for diverse populations.

πŸ›  Tech Stack

IBM watsonx.ai, Python, Flask, JavaScript, HTML5, CSS3, CSV

πŸ”— View Project

Data Assistant Pro: An Automated Data Science Platform

πŸ—‚οΈ Data Science Completed
Data Science Machine Learning Python Streamlit Data Analytics AutoML Data Cleaning EDA Scikit-learn Plotly Portfolio Project

πŸ“ Overview

An end-to-end web application built with Streamlit to solve one of the most time-consuming pain points in data science: the manual process of data cleaning, analysis, and model building. Data Assistant Pro automates the entire workflow, allowing users to upload a raw dataset and receive a comprehensive exploratory data analysis (EDA), apply interactive data cleaning tools (for missing values, duplicates, and outliers), and automatically train and compare multiple machine learning models for both classification and regression tasks. The system is designed to be a full-stack solution that empowers users to move from messy data to actionable insights and live predictions in minutes.

🌟 Impact

This project acts as a game-changer for solo analysts, students, and small teams by dramatically increasing productivity. It automates hours of manual, repetitive work, allowing data professionals to focus on higher-value tasks like feature engineering and interpreting results. By providing an intuitive interface for complex machine learning tasks, it makes advanced data science accessible and empowers users to make data-driven decisions more efficiently.

πŸ›  Tech Stack

Python, Streamlit, Pandas, NumPy, Scikit-learn, Plotly

πŸ”— View Project

AfriRiseHub

πŸ—‚οΈ Web Development Completed
CMS Youth Empowerment Africa Laravel PHP HTML

πŸ“ Overview

A CMS-powered opportunity hub for African youth β€” job boards, scholarships, and AI tools for growth.

🌟 Impact

Empowering African youth with centralized access to life-changing opportunities

πŸ›  Tech Stack

PHP, Laravel, MySQL, TailwindCSS, JavaScript

πŸ”— View Project

Loan Default Prediction

πŸ—‚οΈ Data Science Completed
LOAN PREDICTION LOGISTRIC REGRESSION

πŸ“ Overview

This project aims to help lending institutions identify high-risk borrowers by predicting the likelihood of loan default using historical data. The project follows a full data science pipeline including data cleaning, EDA, feature engineering, model development, evaluation, and presentation. We use Logistic Regression to classify borrowers as likely to default (1) or repay (0) based on their financial and personal attributes.

🌟 Impact

Logistic Regression to classify borrowers as likely to default (1) or repay (0) based on their financial and personal attributes.; Logistic Regression to classify borrowers as likely to default (1) or repay (0) based on their financial and personal attributes.;

πŸ›  Tech Stack

Python, SEABORN,Pandas, Prophet, Matplotlib

πŸ”— View Project

CitiGuard

πŸ—‚οΈ Data Science Completed
PHP MYSQL NLP

πŸ“ Overview

Tok di Tru. Fix di System. CitiGuard is a citizen-led accountability platform designed for Sierra Leone. It allows people to safely report corruption, police misconduct, abuse of power, and systemic issues β€” all while empowering oversight through transparency, secure access, and data insights.

🌟 Impact

πŸ” Anonymous Reporting with support for evidence (photos, audio, video); πŸ“Š Admin Dashboard with charts and report metrics; πŸ“ PDF/CSV Export with filter options (region, district, date, etc.); πŸ—ΊοΈ Region β†’ District β†’ Chiefdom location mapping; πŸ‘₯ Role-Based Access (Super Admin, Admin, Moderator); πŸ“œ Audit Logs, Soft Deletes, and Admin Action Tracking; πŸ”Ž Searchable Admin Panel with secure role management

πŸ›  Tech Stack

Python, Pandas, Prophet, Matplotlib

πŸ”— View Project

AI-Powered Study Companion

πŸ—‚οΈ AI Tools Pending
GPT-4 NLP Tailwind Next.js

πŸ“ Overview

Built an AI chatbot tutor that adapts to a student's learning pace using natural language understanding.

🌟 Impact

Helped 50+ international students prep for exams interactively.

πŸ›  Tech Stack

OpenAI API, Node.js, Supabase, React

πŸ”— View Project

Smart Health Dashboard

πŸ—‚οΈ Health Analytics In Progress
Streamlit Logistic Regression Healthcare Analytics

πŸ“ Overview

Built a real-time data visualization dashboard for predicting patient readmission rates in hospitals.

🌟 Impact

Aimed to reduce readmissions by providing actionable insights to physicians.

πŸ›  Tech Stack

Python, Pandas, Seaborn, Streamlit

πŸ”— View Project

Accident Hotspot Prediction

πŸ—‚οΈ Data Science Completed
Machine Learning Accident Forecasting KMeans Prophet

πŸ“ Overview

Predicted accident hotspots using geospatial clustering and time-series forecasting.

🌟 Impact

Helped New York City visualize critical crash zones and suggest interventions.

πŸ›  Tech Stack

Python, Pandas, Prophet, Matplotlib

πŸ”— View Project