Bhavya Kumar Mishra

Building AI copilots, frontend and backend services for Organization management system in Microsoft's Volume Licensing applications handling 80+ billion dollars of yearly revenue.
Build 0->1 UI & backend solution for managing, searching organizations and contacts, leveraging SQL databases, Azure search, advanced telemetry & analytics, eventually scaling the application to 100k active users, averaging ~1M request / day.
Developed an intelligent classification system using Open AI, clustering techniques, and advanced machine learning algorithms to categorize 1M+ customers, on similarity basis.
Building Agentic AI tools to improve developer’s productivity using Generative AI across 150+ FTEs org-wide that streamlined development workflows and boosted feature rollouts by 40% in six months.
Led the team of 4 developers for development of update statement feature for partner users, resulting in a significant annual reduction in BPO costs and a decrease of 29,000 annual orders that required manual intervention, increasing adoption rate by 42%.
Designed a optimized and robust trade screening transformation system for organizations in VL domain, resulting in meeting the compliance standards and saving 81k hours of potential effort of trade team operational activities.
Led organization-wide security wave initiatives (10+ teams), driving the successful completion of critical security milestones, significantly hardening system.
Engineered a multithreaded C# application to update 1.6M+ customer organization records, incorporating scalable rate limiting and batch processing.

Designed & implemented end to end solution for ETL data ingestion flow and warehousing in PA Intel, providing product keys related information from key production to activation & distribution, consumed by Microsoft Windows, Office, Anti-piracy, Visual studio for business continuity.
Developed & designed end to end Reporting framework solution for product key life cycle involving petabytes-scale data.
Strengthened the security and vulnerability of PRSS by migrating services to secure Azure tenants preventing potential cyber-attack, and creating extensions, automation tools, and QoS dashboards that are utilized by multiple teams within Microsoft.

Modified Burn Rate Calculator using Machine Learning and Anomaly detection- Implemented and compared various ML models to calculate burn rate based on historic time series data. Achieved a significant increase in accuracy with overall 90% plus prediction rate.
ICM Bot Automation Tool - Developed a chatbot for automation of services on servers using Bot Framework, C#, LUIS & PowerShell that can accept a user input in natural language and perform the activity, reducing the overhead of logging in to the server and performing the desired job.