Computer Engineering student with hands-on experience in RAG systems, CNNs, and on-device ML with TFLite. Production Flutter background — 10K+ daily transactions. Eager to contribute to AI-focused teams.
I'm a Computer Engineering student from Surat who blurs the line between mobile engineering and AI research. Two years ago I was shipping production Flutter apps. Today I'm building RAG pipelines, training CNNs, and deploying models on-device — sometimes all in the same project.
Real production exposure means I don't just prototype — I understand scale, reliability, and the cost of mistakes. I bring that discipline into every ML system I build.
Intelligent Q&A system using Retrieval-Augmented Generation for accurate, context-aware responses. Integrated Whisper API for audio transcription, vector embeddings for semantic retrieval, and applied prompt engineering to minimise hallucination.
CNN achieving >98% accuracy on handwritten digit recognition using TensorFlow & Keras. Converted to TFLite and deployed on-device inside a Flutter app — real-time inference, zero internet required.
Supervised ML pipeline for real-estate valuation using regression algorithms. Feature engineering via Scikit-learn's ColumnTransformer; model persisted with Joblib for production deployment.
Regression model predicting BMW prices from model, year, mileage, and fuel type. Full pipeline with data cleaning, outlier handling, and feature encoding for maximum accuracy.