
Learning Engineer Agent
A multi-agent system that drafts eLearning courses from web research and live Socratic interviews with subject-matter experts — built to find where human judgment stays in the loop.
I've led learning and AI products used by millions: AI that answered 85% of 4M+ student questions at Toppr, a zero-to-one product worth $10M+ at Carnegie Learning, and a bootstrapped company I grew from 3 people to 30. Trained in ML research and in HCI at Carnegie Mellon, I still design the product and write the code myself.

A multi-agent system that drafts eLearning courses from web research and live Socratic interviews with subject-matter experts — built to find where human judgment stays in the loop.

The exam-prep company I founded in 2013 and grew from 3 to 30 people, now an AI product: UPSC Mains answers evaluated across 23 papers, with structured feedback in under a minute. On the web, iOS and Android.

Short-form video turned into interactive language lessons, grounded in social learning theory.

A visual tool that lets educators build scenario-based learning activities without code.
Research-led strategy for a 70-branch regional bank; digital adoption up 15%.
Machine learning on satellite imagery and histology at IIIT Hyderabad: five papers, and India's first team at CanSat, a NASA-sponsored satellite competition. Then taught 1,500+ undergraduates at LNMIIT.
Bootstrapped JeetoBharat, an exam-prep company, from three people to thirty: $500K ARR in year one and 10,000+ learners across four states.
AI answers for 4M+ student questions at Toppr, a $10M+ product from zero at Carnegie Learning, and now $10M+ in annual training revenue impact as Principal Product Manager at Axway.
A 10-agent course-design system, an AI examiner for exam answers and a language app in 15 languages, built end to end on Python, Flutter and GCP. Every recommendation comes with a working prototype.
A deep reflection on Khan Academy's mission, international strategy, and the gap between ambitious goals and ground realities in resource-constrained classrooms. Drawing from my interview experience and secondary research across India, Pakistan, Vietnam, and the Philippines.
A founder-style diary on building an agentic EdTech AI platform, tackling web research reliability issues, and applying STORM-inspired thinking.
Great products are born when visionary ideas meet the right technology—and both are executed with precision. It’s not a battle between idea and tech; it’s a partnership where each fuels the other. Success lies in aligning feasibility, viability, and trust into a seamless user experience.