Machine Learning Engineer Intern
• Engineered a diverse set of FastAPI endpoints, enabling seamless creation of customizable Large Language Models (LLMs) and Data Loss Prevention solutions• Integrated TRL, Peft, BitsandBytes, PyTorch, Llama.cpp, and Ollama into a cohesive framework for developing and deploying a robust, high-performance finetuning API endpoint• Leveraged Langchain for document preprocessing and multi-query retrieval, and ChromaDB for optimized storage of document vectors, to enable a Retrieval Augmentation Generation API endpoint• Developed a Data Loss Prevention API by integrating regex patterns into a spaCy pipeline to accurately identify PII and furthermore training spaCy’s Named Entity Recognition for enhanced PII detection• Utilized Langgraph to create Gemma2 based agentic workflows that send SIEM alerts, emails, and texts based on the severity and context of the data leakage when PII is detected