Apoorva Tomar

About Apoorva Tomar

I’m a passionate Generative AI developer with nearly a year of hands-on experience working with LLMs, prompt engineering, and multi-agent frameworks. At Info Origin, I served as an Associate Python Developer, where I helped build large-scale, real-world agentic workflow systems using the LlamaIndex framework. My work focused on designing intelligent, modular agent systems that leverage orchestrator agents to manage and coordinate multi-step tasks. I have extensive experience creating custom agents tailored to specific roles such as information retrieval, summarization, and critique, enabling seamless collaboration within multi-agent setups. I’m also skilled in advanced prompting strategies, including chain-of-thought, ReAct, and role-based prompting—methods that optimize LLM performance, reasoning, and reliability across diverse applications. In addition to my AI expertise, I have a solid foundation in backend development using FastAPI, where I’ve built scalable APIs to serve AI models, process documents, and deliver structured outputs for production use. This combination of AI and backend engineering allows me to deliver full-stack intelligent systems that are both powerful and practical. My technical stack includes Python, LlamaIndex, OpenAI APIs, FastAPI, Docker, and vector databases like FAISS. I understand how to take AI prototypes and turn them into production-ready services that solve real business problems. As I continue to grow, I’m excited by opportunities to work on retrieval-augmented generation (RAG), multi-agent orchestration, and domain-specific AI applications in areas like healthcare, finance, and legal tech. I thrive at the intersection of AI research and practical implementation—and I’m driven by building systems that don’t just generate outputs, but generate real impact.

Experience

1. Associate Python Developer

   Info Origin Inc.

   Sep 2024 — Present · Noida, IN

   Developed and deployed a real-world, large-scale multi-agent workflow system using LLMs and the Llama Index framework, enabling intelligent task delegation and coordination across agents- Designed and implemented custom agents with specialized roles such as data retrieval, summarization, critique, and document processing, integrated into an orchestrator-managed agent pipeline- Applied advanced prompt engineering techniques (e.g, chain-of-thought, ReAct, role prompting) to fine-tune agent behavior and improve the accuracy and consistency of LLM-generated outputs- Built robust backend REST APIs using FastAPI to support AI-powered services, including document upload, parsing, and dynamic interaction with agent systems- Collaborated with cross-functional teams to transform AI prototypes into production-ready solutions, maintaining modularity, scalability, and clear documentation for seamless deployment.

Education

• Dr. A.P.J. Abdul Kalam Technical University

  Bachelor of Technology - BTech, Information Technology