Most AI projects fail not because the models are bad,
but because the problem wasn’t clear, the data wasn’t ready,
or the solution didn’t fit real workflows.

That’s the gap I enjoy working in.

I’m a Data Science graduate student at UMBC working at the intersection of
data, machine learning, and applied AI systems.

I focus on:

• Turning messy data into reliable signals
• Building simple ML/AI systems that scale
• Translating analysis into decisions people can act on

No hype. No overengineering. Just systems that work.

• Start with the problem, not the model
• Data quality > model complexity
• Systems beat tools
• If it can’t be explained simply, it’s not done

Languages
Python · SQL

ML & AI
Scikit-Learn · PyTorch · TensorFlow · XGBoost
NLP · LLMs · Time-Series · Classification

Data & Analytics
Pandas · NumPy · EDA · Feature Engineering

Visualization & BI
Power BI · Tableau · Matplotlib · Seaborn

Systems & Platforms
AWS · LangChain · Streamlit · Zapier · Git

🔹 AI Tutoring Session Summarizer
LLM-powered pipeline that automatically summarizes sessions and delivers insights via email.

🔹 Crime Analytics & Prediction Chatbot
ML-driven chatbot combining crime, weather, and demographic data for predictive insights.

🔹 Forecasting & Classification Systems
End-to-end ML workflows focused on interpretability and real-world performance.

(Scroll down to explore the repos 👇)

• Agentic AI & workflow automation
• RAG-based analytics systems
• ML deployment patterns that don’t break in production
• AI systems for small & mid-sized businesses

I’m always interested in:

• Applied AI / ML projects
• Data systems with real-world impact
• Collaborations that value clarity over complexity

📫 Let’s connect on LinkedIn
⭐ Star a repo if something here helps you

“The goal isn’t smarter AI.
The goal is better decisions.”