Hi, I'm Shreya. Welcome!

I've spent the last 6 years at the intersection of AI, product strategy, and data—building systems that turn complex problems into reliable, scalable solutions. What started as a curiosity about how data could drive decisions has evolved into a deep focus on AI-native products and the frameworks that make them trustworthy.

At Travelopia, I've led AI-first product delivery across luxury travel brands, engineering privacy-compliant AI systems and transforming how teams work. I've reduced cycle times by 80%, boosted sprint predictability from 55% to 95%, and delivered secure payment integrations processing $760K in 12 days—all while keeping security and user experience at the center. Before that, at Signeasy, I used data to improve retention by 10% and influenced pricing strategy through insights that changed how the company thought about product prioritization.

What drives me is the gap between what AI promises and what it actually delivers in production. I build evaluation frameworks, run failure analysis, and create systems that keep AI, delivery, and teams aligned. I've designed 5-factor evaluation rubrics that achieved 98% pass rates, automated workflows that saved weeks of manual work, and built tools that match receipts to bank statements using GPT-based fuzzy logic—because real product work means solving problems that matter, not just shipping features.

My approach blends structured thinking with fast experimentation. I care deeply about making things clearer, faster, and more reliable—whether that's a product backlog, an AI evaluation framework, or a team's delivery rhythm. I've supported $300M+ revenue portfolios, led agile transformations, and mentored teams because I believe great products come from great systems, and great systems come from people who think deeply about how things actually work.

Outside work, I'm constantly learning about evals, tinkering with projects, failing a lot, and questioning everything. I write about AI and product on LinkedIn, connect with like-minded people, and spend my evenings breaking models on purpose—because understanding how things fail is how you build things that don't.