Apr 2020 – Jan 2023 · 2 yrs 9 mos · India

• Launched NewzGPT, an early LLM system for real-time short-video news.
• Grew to 20k+ MAUs, bridging factual accuracy and entertaining delivery.
• Showcased how specialized AI can reshape content consumption at scale.

Oct 2017 – Sep 2019 · 1 yr 11 mos · San Francisco, California

• Led recommendation projects powering ~40% of user engagement.
• Delivered a 7.1% CTR lift and $17.5M revenue gain via advanced ML ranking.
• Scaled systems to 175M WAUs, handling 65k QPS with robust AI pipelines.

Jul 2017 – Oct 2017 · 3 mos · San Francisco, California

• Joined post-acquisition of Post Intelligence, built early foundations of AWS Honeycode.
• Focused on no-code application building, a precursor to automated workflows.
• Contributed to bridging dev-tool complexity for mainstream business users.

Oct 2015 – Jul 2017 · 1 yr 9 mos · San Francisco, California

• Invented a patented Bayesian ads system (US15/427550), improving RPM up to 15%.
• Built a deep learning engine predicting tweet virality & personalizing content suggestions.
• Key driver of the startup’s success and eventual acquisition by AWS.

Oct 2014 – Oct 2015 · 1 yr · Madison, Wisconsin Area

• Developed orthopedics modules, blending user-friendly widgets with robust clinical data.
• Completed 6-month advanced training in M, VB6, Hyperspace—enabling faster feature rollouts.
• Improved workflows for thousands of clinicians using EpicCare Ambulatory.

Jun 2013 – Jul 2013 · 1 mo · Greater Bengaluru Area

• Developed a novel news recommender using Twitter signals + Yahoo! News, boosting user engagement.
• Pioneered an algorithm ranking macro categories via Wikipedia/Freebase, elevating relevance.
• Awarded Best Intern (1/20) for advanced ML-driven content personalization—sparked my future in AI.

May 2012 – Jul 2012 · 2 mos · Ecole Polytechnique Fédérale de Lausanne, Switzerland

• Built privacy-preserving developer tools on the Qt framework, pioneering user-centric protection.
• Authored a runtime “privacy score” simulator, influencing app design for minimal data intrusion.
• One of six students chosen from India—solidified my drive for cutting-edge, user-focused innovation.

May 2011 – Jul 2011 · 2 mos · Center for Development of Advanced Computing, Pune

• Enhanced HPC scheduling by utilizing idle CPU “pockets”—improved cluster efficiency to 98%.
• Innovated a predictive approach to schedule auxiliary tasks without abort/purge, cutting wasted cycles.
• Bolstered my belief in algorithmic optimization—a precursor to building scalable AI solutions.