Machine Learning Engineer @ Microsoft | Responsible AI
I am a Machine Learning Engineer passionate about building scalable, modular AI solutions that drive meaningful change. Grounded in engineering fundamentals, I design and develop cutting-edge AI technologies that evolve with today’s rapidly changing landscape.
Detecting AI safety and security risks in Microsoft AI products
•
Research and productization of novel AI defense techniques (see - https://arxiv.org/abs/2509.05608)
•
Lead product teams through the company-wide Gen AI RAI compliance processes
AI Product
Advancing agentic capabilities for Microsoft Copilot Studio (https://www.microsoft.com/en-us/power-platform/blog/power-apps/new-smart-paste-makes-filling-forms-as-easy-as-copy-paste/)
Microsoft AI Development Acceleration Program (MAIDAP), https://www.microsoftnewengland.com/maidap/
Spring 2024: Office of the CTO (OCTO)
•
Developed open-source Semantic Kernel agent framework: https://github.com/microsoft/semantic-kernel/commit/ce6ccd7e86be3a8130f5cf9724fa5e22fe17c3f1
•
Filed three patents following the research findings
Fall 2023: Microsoft Digital (MSD)
•
Scoped and implemented a new AI Plugin for M365 Chat that is projected to alleviate up to 80% of the benefits support queue by streamlining answer-finding and discovery of new HR benefits.
•
Leveraged RAG architecture, ACS, Cognitive Services, and OpenAI services
Spring 2023: Azure Machine Learning and Microsoft Research https://techcommunity.microsoft.com/blog/machinelearningblog/debug-object-detection-models-with-the-responsible-ai-dashboard/3825658
•
Added full stack support for object detection scenario within Responsible AI Dashboard, i.e. programmatic generation of explainability visualizations.
•
Released the feature to public preview at Microsoft Build
•
Proactively sought feedback from external customers, feature was estimated to reduce development time by >50%
Fall 2022: Business Applications Platform (BAP)
•
Designed and implemented a new ML pipeline for a Smart News feed using AI Builder (AzureML), which simplified the onboarding process for new customers (saving 6 weeks of engineering cycles per customer)
•
Coordinated data and integrated responses from 3 new APIs: Bing, Cognitive Services Named Entity Recognition, and Module Execution Framework.
