Optimized prompt engineering and RAG logic, deployed with Docker and CI/CD on AWS Lambda, ensuring scalability and reliability.

Developed a full-stack fitness platform with FastAPI REST API and React Native mobile app. Integrated LLM APIs for personalized guidance and implemented gamification.

Developed a high-precision RAG pipeline, intelligent chunking, Metadata Boosting, and containerized deployment using Docker. Ensured Clean Architecture and SOLID principles.

Designed a hybrid search engine with LangChain, vector embeddings, and RRF algorithm, improving precision. Built real-time LLM streaming with SSE and a modern React frontend.

Handled data cleaning, EDA, normalization, feature engineering, and documented the workflow for traceability.

Designed a predictive model for crop yield with R² ≈ 0.92 and RMSE ≈ 1.38 t/ha. Built a full data pipeline and ensured model reproducibility.