PRATIK RAJ

I’m currently pursuing a B.Tech in Data Science (Class of 2027), and I’ve spent the last couple of years building real-world data projects that taught me how to turn raw numbers into useful business insights. I enjoy cleaning messy datasets, writing efficient SQL queries, and building dashboards that help people make decisions.

In my Retail Sales BI Dashboard project, I loaded a 10,000+ row dataset into MySQL, created a clean Python pipeline with SQLAlchemy, and used advanced SQL (window functions and CTEs) to identify monthly trends and best-performing products. I also built a Streamlit dashboard with filters so non-technical team members could explore key metrics, and I used outlier analysis (IQR & Z-score) to uncover that orders with discounts over 40% were almost always making a loss. Beyond BI, I developed a fully local AI chat interface using FastAPI and Ollama that answers questions from uploaded PDFs/CSVs without relying on paid APIs, and I’ve strengthened my foundation through Oracle Generative AI and Hugging Face NLP training, plus open-source contributions during Hacktoberfest 2024.