Rajvardhan Pawar

AI Engineer with 2+ years of experience designing and deploying Generative AI systems in production including Agentic RAG pipelines, LLM fine-tuning workflows, and multi-agent orchestration frameworks. Proficient in LangChain, LangGraph, Hugging Face Transformers, and open-source LLMs (Mistral, LLaMA 3), with a strong backend foundation in FastAPI, Celery, Redis, PostgreSQL, and Docker. Focused on building reliable, scalable, and intelligent systems that solve real business problems end-to-end.

Software Engineer at IMR Tech Solutions (2025-02 – Present)

Architecting and deploying Generative AI systems for market research automation and LLM fine-tuning workflows.

• Architected a multi-stage Generative AI pipeline for automated market research using Perplexity AI and OpenAI, reducing report turnaround from days to minutes and eliminating manual research effort entirely.
• Fine-tuned Mistral-7B on domain-specific corpora using LoRA (PEFT) via Hugging Face Transformers, improving output relevance and reducing hallucination rate for structured report generation tasks.
• Built an Agentic RAG system with LangChain and ChromaDB applying Prompt Context Engineering (chain-of-thought, few-shot prompting) to maximize retrieval quality and LLM output accuracy.
• Designed distributed async task execution using Celery and Redis for concurrent LLM inference, improving system throughput by 80%; optimized PostgreSQL queries and Redis caching reducing API latency by 35%.
• Deployed containerized AI services via Docker-based CI/CD pipelines; built PyTest suites, participated in 40+ code reviews, and used GitHub Copilot as a core daily development tool.

Junior Software Developer at Walstar Technologies (2024-01 – 2024-12)

Backend and frontend development with focus on API design, code quality, and component architecture.

• Developed backend APIs in Python and Node.js with authentication, validation, structured error handling, and secure coding practices.
• Contributed to Agile sprint code reviews, feature planning, feasibility analysis, and technical solution evaluation.
• Built reusable React and TypeScript components following component-based architecture, reducing code duplication by 25%.

Bachelor of Technology in Computer Science and Engineering – KIT's College of Engineering (2019-07 – 2023-07)