Divya Jaisansaria

I've built applied AI products across Wells Fargo and Genus Power Infrastructures, from production frontend improvements to multilingual support systems. At Wells Fargo, I closed 7+ Jira stories, delivered 20+ frontend features and fixes, and created 100+ Playwright E2E tests across 40+ critical user journeys.

At Genus Power Infrastructures, I designed a multilingual AI support agent with LangChain, Groq APIs, and a multimodal RAG pipeline that resolved queries across 1,000+ document chunks with PDF page redirection and persistent chat history. I also built React.js dashboards and REST API workflows that reduced manual operations by 50%.

My research on EEG-based depression detection uses a hybrid CNN-BiLSTM architecture that achieved 96.2% accuracy and a 0.95 F1-score on the MODMA EEG dataset, and has been accepted at ComSIA Conference 2026.

I'm also building practical AI tools such as Udaan, an assistant for rural women that turns voice and image inputs into e-commerce listings, and Connecting the Dots, a multimodal RAG document-intelligence platform. I enjoy taking ideas from model development through usable, automated product experiences.