Looking for a job

RISHIT AGARWAL

@rishitagarwal

Message

I build production GenAI, RAG, and computer-vision systems for insurance automation.

India

Message

What I'm looking for

I'm looking to build production GenAI, RAG, NLP, and computer-vision products that automate complex workflows and improve customer experiences, while continuing to deepen my work in LLM reasoning and applied AI research.

At PolicyBazaar, I build production GenAI, RAG, NLP, and computer-vision systems that automate insurance workflows and improve customer self-service.

I trained a YOLO-based Aadhaar/KYC detection model to 99.2% accuracy for real-time PII masking, and built a motor-policy RAG assistant using document parsing and agentic AI.

I’ve also built health-insurance recommendation and advisor chatbots, a cross-sell intent platform from sales interactions, and an NLP compliance checker for advertisements and ASCI rules.

Alongside industry work, I research LLM reasoning at the University of South Carolina, studying graph-based prompting, constraint-satisfaction retrieval, and zero- and few-shot generalization.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist | GenAI | RAG | LangGraph

Current

PolicyBazaar

May 2025 - Present (1 year 4 months)

Developed a YOLO-based model for automated PII masking in KYC documents, achieving 99.2% accuracy. Built production RAG pipelines and AI chatbots for customer self-service and internal advisor support.

YOLO PyTorch OpenCV fastAPI Docker Lang Chain LangGraph DocLayout Gemini Vision OCR Uvicorn Python Pandas NumPy AWSAthena BERT NLP Prompt Engineering ChatGpt Claude

Current

Junior Research Intern | LLM Reasoning

Current

University of South Carolina

Jun 2024 - Present (2 years 3 months)

Studied constraint-satisfaction retrieval using state-of-the-art LLMs through graph-based prompting. Benchmarked instruction-tuned LLMs on arithmetic, grammatical, and commonsense reasoning tasks.

Reasoning Graph Based Prompting React Few Shot Learning Benchmarking Instruction Tuning

Data Scientist Intern

PolicyBazaar

May 2024 - Jul 2024 (2 months)

Built automated KYC key-value extraction using Table Transformer, Azure OCR, and LayoutLM. Developed a 10+ class document classification system using text, spatial, and vision embeddings.

Table Transformer Azure OCR LayoutLM Document Classification Text Embeddings Vision Embeddings

Product Management Intern

Logy.AI

Sep 2023 - Mar 2024 (6 months)

Built a GenAI chatbot for oral/eye screening and designed the conversational context-management pipeline. Mapped the cancer-patient journey to inform a GenAI companion supporting patients through treatment.

GenAI Chatbot Development Conversational AI Context Management Product Management