About

Tanay Pratap is an applied intelligence leader with over a decade of experience building systems at the intersection of AI, large-scale software, and institutional problem-solving. His career spans enterprise infrastructure, deep learning research, product engineering at Microsoft, venture-backed startups, and applied AI research — a trajectory that has consistently moved from building technology to deploying it where it creates measurable impact.

Research & Current Work

Today, Pratap leads On Ground Labs, an applied AI research initiative built on a simple thesis: AI research is too abstract and too distant from everyday realities. On Ground Labs builds systems that stay close to the world they serve — human-first, empirically grounded, and designed for deployment, not just publication.

His current research spans small language models designed for edge deployment, improved retrieval architectures, and AI alignment. He has published multiple research papers, filed patents across these domains, and authored a textbook on LLM training that is currently part of the curriculum at Manipal University's MTech programme.

In parallel, he consults with enterprises — working with US banking institutions on intelligent document processing and agentic automation, and with healthcare providers in the Middle East on AI systems for patient care, diagnostics, and medical coding.

Building Companies

After Microsoft, Pratap founded and led Invact Metaversity, raising $5 million in venture funding to build a VR and WebXR education platform — one of India's earliest experiments in immersive technology for learning at scale. He built and led teams across product, engineering, and research functions.

Alumni of his programmes have gone on to roles at Google, Uber, GitHub, Flipkart, and Razorpay, among others — a testament to the rigour of the technical talent pipeline he built.

The Ecosystem

Three ventures, one thread: applied intelligence. Microsoft taught me how to build products at scale — 100M+ users on Teams, 70M on Outlook PWA. Invact applied that to education — a research-driven VR learning platform, $5M raised, 100+ alumni placed. On Ground Labs pushes the frontier — 23 researchers running 14 projects in small models, benchmarks, and agentic systems.

Technical Foundation

Pratap began his career in 2012 at Cypress Semiconductor, where he built and managed network infrastructure across the APAC region. He updated operational best practices, trained the NOC team for ongoing upkeep, and worked directly with vendors including Cisco, Aruba, Airtel, and Azure to negotiate pricing and ensure support quality — an early education in how large institutions operate at the infrastructure level.

His work in artificial intelligence began in 2015 at Cisco, where he built NLP systems to classify and diagnose faults across massive router log datasets. In 2017, he joined an early-stage startup as a founding engineer, applying computer vision to help consumer brands improve sales performance.

In 2017, he joined Microsoft, where he spent four years building products at scale. He was part of the team that grew Microsoft Teams from a new product to a platform serving over 100 million users, and later built the Progressive Web App for Outlook, serving 70 million users globally. This period shaped his understanding of what it takes to ship reliable, high-performance software to hundreds of millions of people.