At LIGHTING BEES PRIVATE LIMITED, I developed AI-powered accounting and filing software in Appsmith, including an invoice processing system that automated extraction, classification, and validation.
I implemented document processing pipelines that sent individual invoices to prompted Gemini models for data extraction, then structured the results in database tables with frontend validation workflows. I also added batch processing and scheduler-based execution to optimize backend operating costs.
At Naest Tutors, I developed backend services for a home tutoring platform, building REST APIs with Node.js and Express.js and implementing JWT authentication and role-based access control.
At BR SoftSol, I built data analysis workflows and implemented machine learning algorithms. My projects include a Flask-based disease prediction application and a crime behavior analysis project using time-series forecasting; I also published research on CNN models for eye disease classification using retinal images.
