About
Computer Science (Data Science) graduate with hands-on experience in AI data annotation workflows, NLP, speech-to-text technologies, LLM evaluation, data validation, and quality assurance. He has worked on structured evaluation, model output review, and improving data quality for GenAI and NLP workflows.
Experience
-
AI/ML Developer InternFestiva MomentsFeb 2026 – Jun 2026
Supported data annotation and labeling workflows for GenAI and NLP model development. Improved held-out task accuracy by 28% on a domain-specific LLM using Hugging Face Transformers and PyTorch, reduced data-quality incidents by 38% through automated LLM evaluation scripts, built and evaluated a RAG pipeline over 10,000+ business records, and performed structured evaluation across model and prompt variants in MLflow.
-
Python Full Stack Developer InternBrain O Vision Solutions Pvt LtdMar 2024 – Jun 2024
Implemented automated data validation checks, optimized T-SQL and ANSI SQL pipelines on SQL Server, developed REST API integrations and reporting workflows, and created Power BI dashboards while reviewing structured data for reporting accuracy and consistency.
Education
-
B.E.AMC Engineering CollegeComputer Science & Engineering (Data Science) · Dec 2022 – Jun 2026
Skills
Tools / apps / platforms
Languages
Projects
-
Build-Your-Own-GPT & Prompt EngineeringPython, PyTorch, Hugging Face Transformers
Implemented a transformer training and fine-tuning workflow with tokenization, sampling controls, model evaluation, and human-response checks.
-
Intelligent Internet Search EnginePython, FAISS, NLP, Sentence Transformers
Built a retrieval and ranking workflow combining lexical search, dense embeddings, and cross-encoder reranking, and evaluated results through structured testing and relevance analysis.
-
Agentic AI SystemPython, LangGraph, LangChain, FAISS, FastAPI
Built a multi-step AI system using tool calling, retrieval, persistent memory, and structured output validation with fallback mechanisms to reduce hallucination rate.
-
Generative AI RAG ChatbotPython, LangChain, FAISS, Pinecone, Streamlit
Built a document question-answering system over PDF and web content and evaluated generated responses against a ground-truth QA set using structured criteria.