At Injala, I build Python backend services and REST APIs for two production fintech platforms: Asuretify, for insurance document intelligence, and Wrap, an insurance rating engine.
On Asuretify, I contributed to a multi-stage pipeline that extracts information from ACORD and endorsement documents using LayoutLM, YOLO, OpenAI APIs, and RAG. I also refined test cases and data-validation checks to improve pipeline accuracy and consistency.
For Wrap, I developed backend services for an LLM-powered rating engine and integrated an AI-based validation gate to accept or discard generated responses before output. I monitored accuracy and response-time metrics across development and staging to support reliability.
I’ve also worked on LLM fine-tuning for document classification, serving a model with FastAPI and Docker and deploying it to Azure. My projects include Atlas, an agentic AI system with PostgreSQL-backed multi-user persistence, and DevContext, a set of Dockerized services orchestrated by a LangGraph agent.
