BS in Data Science and Applications, IIT Madras (Expected 2025)
Hello there! 👋
I’m Utkarsh Shukla, a data science enthusiast with a knack for solving real-world problems through innovative solutions. My expertise lies in machine learning, data preprocessing, and model deployment. I thrive in environments where I can combine my analytical mindset with a creative approach to problem-solving.
- Programming Languages: Python, SQL, Java, Bash
- Data Science Tools: NumPy, Pandas, Scikit-learn, LightGBM, SARIMA
- Web Development: Streamlit, Backend Development
- Mathematical Foundations: Calculus, Linear Algebra, Graph Theory
- Professional Skills: Project Management, Team Collaboration, Communication
A Retrieval-Augmented Generation (RAG) based system for querying PDFs using local LLMs.
- Privacy-Focused: Runs fully offline using local LLMs like DeepSeek.
- Efficient Retrieval: Uses vector embeddings for context-aware document search.
- Streamlit UI: User-friendly interface for seamless interaction.
A content-based filtering movie recommendation system built using Streamlit and TMDB API. Features include:
- Cosine Similarity algorithm for ranking recommendations.
- Interactive UI with dynamic movie poster displays.
A real-world business optimization project focusing on:
- SARIMA-based forecasting for improved sales accuracy.
- Distributor segmentation via K-means clustering.
- Expiry management through an early warning system.
Achieved 92.17% accuracy using LightGBM for predicting mental health risks.
- Interactive app for healthcare professionals via Streamlit.
- Advanced feature engineering for identifying key risk factors.
Predicts campaign success rates using ensemble models (Logistic Regression, Random Forest, etc.), with an F1-score of 0.70178.
- Diploma in Data Science (Dec 2024)
- Machine Learning using NumPy (Sep 2024)
- Academic Mentor Certification (IIT Madras)
Feel free to reach out for collaborations, discussions, or just to say hi:
📧 Email: 21f2001497@ds.study.iitm.ac.in
📱 Phone: +91-6388807839