Atul Singh

Senior GenAI engineer

CoFoundersLab • Tiempo completo

May 2023 - Present • 3 yrs 4 mos

Senior Software Engineer | Coforge Limited | Since May 2023 I work on GenAI and LLM-based solutions for enterprise clients in the fintech domain, designing and building production-ready AI applications from the ground up. Key Projects & Achievements: RAG-Based Document Q&A Systems Designed and built retrieval-augmented generation chatbots that let users "chat with" large document sets. Implemented semantic search using vector embeddings (FAISS) combined with LangChain and OpenAI/Gemini models, delivering accurate, context-aware answers instead of generic chatbot responses. Multimodal Document Retrieval Built a hybrid text + image retrieval pipeline for image-heavy technical documents, combining semantic embeddings with re-ranking strategies. Evaluated multiple vector database and retrieval-fusion approaches to ensure the solution could scale for enterprise-level deployment. AI-Powered Legacy Code Modernization Designed a 3-agent LLM pipeline (Analyzer, Converter, Validator) to convert legacy COBOL/Java codebases into clean, type-hinted Python 3. Built a self-healing validation loop using automated testing (pytest) so converted code is verified, not just translated — reducing manual QA effort significantly. AI Resume Screening Tool Built an end-to-end resume analysis tool using LangChain, OpenAI GPT-4, and PyPDF2, with a FastAPI backend and Streamlit interface — automating what would otherwise be manual resume screening. Cross-Functional Collaboration Regularly collaborate with backend engineers, business analysts, and QA teams to ship AI features end-to-end. Also mentor junior team members on RAG and LLM best practices, helping them ramp up on GenAI development. Core Strengths: Turning complex AI concepts into working, production-ready tools Clean, testable code (FastAPI, pytest, type-hinted Python) Clear technical communication — I explain what I build in plain language, not just jargon