At 6sense, I architected and scaled an enterprise ML platform supporting 150K+ trained models, 11K+ production models, and 20K+ daily inference jobs. I improved training success from under 80% to over 90% and reduced compute consumption by an estimated 33%.
I also designed production model explainability systems and an MCP-based architecture that lets AI agents retrieve model insights and investigate failures. Before 6sense, I designed a Corporate Actions module at Saxo Bank and developed document-processing services at Sage.
