Aspiring Data Scientist / Machine Learning Engineer with strong knowledge in Probability & Statistics, Machine Learning, Deep Neural Networks, and Gen AI, including hands on experience with LLMs, LangChain and LangGraph for building intelligent applications.
1. Customer conversational intelligence platform powered by an LLM agent, 03/01/25 to 05/01/25
1. Developed an end-to-end conversational AI pipeline using NLP, LLMs, and machine learning for sentiment analysis, intent recognition, and topic modeling on customer support conversations
2. Implemented GenAI techniques, including prompt engineering, RAG, function calling, fine-tuning (DistilBERT), and LangGraph architecture with transformer models
3. Fine-tuned DistilBERT on a custom Twitter dataset and applied zero-shot classification to identify user intents and categorize queries into predefined support topics using open datasets
4. Achieved 89% sentiment analysis accuracy, 79% intent recognition accuracy, and 84% topic classification accuracy; chatbot evaluated using RAGAS with 89% faithfulness and 77% answer relevancy
2. SmartBot NDS - GenAI-powered smart assistant using Google Gemini and LangGraph - part of the 5-day Google Gen AI Intensive Course, 04/01/25
1. Built a GenAI-based smart assistant leveraging Google Gemini, LangChain, and LangGraph to retrieve, reason, and respond across multiple support use cases
2. Implemented retrieval-augmented generation (RAG) using order and cancellation policy documents to generate context-aware responses, and integrated function calling to handle dynamic, tool-based queries by interacting with customer transaction and demographic datasets
3. Enabled real-time grounding using Google Search and external APIs to fetch and incorporate up-to-date information into responses, and employed prompt engineering techniques to control LLM behavior, especially during grounding and document-based reasoning