From routing protocols on embedded systems to AGI personalization at Amazon to running autonomous AI engineering teams.
View on LinkedIn →Running end-to-end engineering and science teams built entirely on AI agents. They build the application, generate data, train models, interpret results, and feed it all back — a recursive improvement loop.
IC deep-dive into core GenAI. Principal responsible for the "memory pattern" enabling Amazon's AGI models to deeply personalize experiences across businesses.
Secured AWS CEO funding by synthesizing insights from 300+ customer interviews across 45 enterprise accounts. Led 13-person product/design team through 3 major pivots to first Q-apps launch with $100MM+ revenue potential.
Designed value propositions for 3 new areas. Launched strategic integrations with Zapier and Slack. Created the pricing proposal for Honeycode, approved by Andy Jassy.
Founding PM for Azure Backup. Built VM-backup vision from scratch and launched it. Growth hacked the business 10-fold. Landed as leader in Gartner Magic Quadrant.
Prototyped data visualization in medical imaging, aeronautical, and CAD sectors. Exited after negated hypotheses.
Improved quality adherence KPI by 51% while saving $600K with improved workflows for technical support. Owned routing modules, solved long-standing issues improving routing performance by 3.5%.
Multi-modal AI converts demonstrated processes into executable automation
Natural-language spreadsheet generation with per-user permission filtering
Auto-adapting app UIs from spreadsheet data changes
Permanent row addresses that survive reshuffling (2 patents)
Natural language to structured table updates
"Started as a software engineer writing routing protocols. Ended up running teams of AI agents that build software while I sleep."