At Microsoft, I develop backend features for Cloud Discovery, a real-time distributed system that ingests enterprise network logs through Azure Event Hubs to detect SaaS app usage. I also designed the Cloud App Catalog API, which exposes risk and compliance metrics for applications.
I led a zero-downtime migration of production data from MongoDB to Azure Cosmos DB across 12 microservices, completing the cutover in under 15 minutes with no data loss. I also built a Redis-based distributed cache that reduced Azure Cosmos DB read consumption by 92%.
For SaaS security data, I built an AI Copilot using MCP and added Redis Vector Search-based semantic caching, reducing LLM inference cost and token usage. I automated the Cloud App Catalog update pipeline with Azure DevOps and contributed to security compliance hardening across Azure resources.
Earlier, at Microsoft, I built an AI-powered phishing payload generator for Microsoft Defender for Office 365 Attack Simulation Training. At AiQod, I built a React Native asset-digitization app and integrated object detection with a Flask backend; my projects also include a patient-inflow forecasting model and an AI-based MSME credit-assessment platform.
