Tirth Patel

Background

About Tirth Patel

Data scientist by day, AI agent whisperer by… well, also by day. I build intelligent systems that don’t just crunch numbers — they talk back, solve problems, and (occasionally) give lawyers superpowers. At the Indiana Supreme Court, I designed a full-stack Retrieval-Augmented Generation (RAG) agent that lets legal teams query pages of courtroom transcripts like they’re chatting with a supercharged legal intern — one that never takes a coffee break. We’re talking 80–90% reduction in manual review time. No objections there. Before that, at Prediction Guard, I created GenAI-powered sales agents that scaled outreach 3x and dropped lead qualification costs from $50 to less than a nickel. Basically, I automated the part of sales everyone secretly dislikes. You’re welcome, Sales. Previously in real estate analytics, I optimized operations for 600+ tenants, built scenario-based financial models, and boosted asset performance by 30%. Turns out, good dashboards do make you money. Outside of work, I’m obsessed with merging LLMs, clean data pipelines, and prompt engineering into real-world solutions. Whether it’s forecasting NCAA Final Four picks with 63% accuracy (hi, March Madness fanatics) or diagnosing brain tumors with 95% test accuracy using ViT embeddings — I love tackling problems with data, creativity, and zero unnecessary meetings. Currently wrapping up my MS in Business Analytics & Info Management at Purdue, where I also serve as Vice President of the program (read: part-time event planner, full-time spreadsheet wrangler). If you’re working on anything at the intersection of GenAI, real-world data, and automated decision-making — or if you just want to geek out over Transformers (the models, not the robots… unless you’re into both) — let’s connect!