Open to opportunities

kabir rao

@kabirrao

I build scalable AI document-extraction platforms, distributed backend services, and high-throughput automation pipelines.

India

Message

At BNY, I help build and scale an AI document-extraction platform that processes PDFs, scanned documents, CSVs, and Bloomberg screenshot data for enterprise automation. I own a 128K+ document, 36-market workflow and have contributed directly to approximately 20 FTEs of savings.

I build production pipelines using Java, Python, LLMs, vision OCR, microservices, and event-driven systems. My work includes dual-model extraction, multithreaded components, HITL review interfaces, observability with Splunk, Snowflake data integrations, and S3-based audit traceability.

I also shipped Nexora, a live multi-agent document-extraction product built with FastAPI, Next.js, Redis/Arq, Supabase, and LLM-powered extraction workflows.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer (AI Hub)

Current

BNY

Jul 2024 - Present (2 years 2 months)

Core member of the team that built and scaled the Instruction Capture Pipeline, a config-driven, multi-agent document-extraction platform. Engineered dual-model LLM extraction and vision-OCR ingestion, onboarded multiple use cases, and developed distributed microservices for an AI innovation platform.

Java Python JavaScript Spring Boot PostgreSQL Microservices REST APIs LLM OCR Multi Agent Systems Splunk Snowflake Docker Kubernetes

Education

Degrees, certifications, and relevant coursework

Indian Institute of Technology Guwahati

Bachelor of Technology, Mathematics and Computing

2020 - 2024

Pursued a Bachelor of Technology in Mathematics and Computing, gaining a strong foundation in both theoretical and applied aspects of computing and mathematics.