- Built robust data pipelines to clean, process, and load data for Retrieval-Augmented Generation (RAG) systems, ensuring efficiency and reliability.- Fine-tuned and evaluated models for tasks like text classification, sentiment analysis, and speech-to-text, delivering high-performance results.- Explored and implemented cutting-edge approaches for knowledge bases, using graph-enhanced RAG to improve data retrieval and understanding.- Developed innovative methods for reasoning at inference time, improving model accuracy and delivering better predictions.- Created LLM-powered applications integrated with vector databases like Qdrant and added caching to manage chat memory seamlessly.- Established best practices for prompt engineering, tailoring solutions to specific domains for more precise and impactful results.

- Responsible for building LLM applications.- Skilled in conducting statistical analysis using Pandas, NumPy, and Matplotlib.- Developed Python scripts for automation and data scraping, streamlining processes, and enhancing data acquisition.

- Addressed various classification and regression challenges.- Executed 2D to 3D image reconstruction tasks using Nerf.- Provided crucial support in debugging and issue resolution.