Greater Delhi Area
I am a Senior AI Engineer with 5 years of experience across industry and research, including 3 years in industry and 2 years of research at IIT Bhubaneswar. My core areas include (not limited to): - Generative AI and Large Language Models - Retrieval-Augmented Generation - Natural Language Processing - LLM Fine-Tuning with LoRA/QLoRA - AI Agents and Workflow Automation - MLOPs - Python, FastAPI, Docker, Ollama, OpenWebUI - Vector Search, Embeddings, FAISS, and Semantic Retrieval I specialize in Generative AI, LLMs, NLP, RAG pipelines, transformer-based architectures, AI agents, and production-grade machine learning systems. My work focuses on building intelligent AI solutions that move beyond prototypes and create real business impact. At Alchem International, I work on production GenAI systems, and built multiple business use case ai solution, fine-tuned model pipelines, and AI automation workflows. During my research at IIT Bhubaneswar, I worked on LLMs, multimodal AI, text-to-image generation, prompt expansion, fine-tuning, and retrieval systems. I am focused on building reliable, scalable, and high-impact AI systems for real-world enterprise use cases.
- Built and deployed production-grade GenAI systems for enterprise workflows, including natural language query understanding, SQL generation, document intelligence, and spreadsheet analysis. - Designed a domain-specific LLM pipeline for converting business questions into SQL across multiple data tables, improving report generation speed and reducing manual dependency. - Worked on fine-tuning LLMs using LoRA/QLoRA techniques to improve domain-specific SQL generation accuracy and reduce runtime prompt complexity. - Built intelligent Excel and document analysis pipelines combining deterministic computation with LLM reasoning for accurate business data interpretation. - Integrated OpenWebUI, Ollama, Docker, and custom FastAPI router services to support local LLM deployment, model routing, streaming responses, and file attachment processing. - Improved AI system reliability by designing validation layers, schema-aware processing, prompt optimization, and compact context generation for LLM inference. - Led AI development efforts across model selection, backend architecture, deployment debugging, and production workflow improvements.
1. Developed and configured an AI-driven document-based Q&A chatbot using Streamlit, FAISS, and OpenAI to provide accurate responses based on PDF inputs. The system was directly used by 2.3 million users during the COVID era. 2. Collaborated with offshore teams, managing end-to-end project development with Agile methodologies, ensuring timely and high-quality deliveries via Git and JIRA. 3. Led technical discussions and requirement analysis, integrating efficient retrieval systems with FAISS, embedding models, and document similarity search for enhanced user experience.