I enjoy working on problems where data, logic, and real systems come together. Most of my time goes into building machine learning–based applications and understanding how models behave once they move beyond notebooks.
I like working end-to-end — from cleaning data and training models to exposing them through APIs and deploying usable systems. I’m currently focused on strengthening my fundamentals while building projects that reflect practical problem-solving.
- Python, C++, SQL
- Machine learning and deep learning
- FastAPI, Streamlit
- Docker, Git, GitHub
- Relational databases and caching systems
-
Insurance Premium Category Predictor An end-to-end ML system predicting premium tiers. I used Docker Compose to manage a FastAPI backend and a Streamlit frontend as decoupled, containerized services.
-
Music Recommendation System A hybrid recommender utilizing the Spotify API. I implemented Redis caching to store metadata, successfully reducing API calls and minimizing request latency.
-
Sign Language Recognition A Deep Learning project using MobileNetV2 for real-time ASL recognition. The focus was on optimizing the trade-off between model accuracy and real-time inference speed.
- 📧 Email: pappumeghana2006@gmail.com
- 🔗 LinkedIn: linkedin.com/in/pappumeghana432