Vinod Joshi

Lead the E2E development and deployment of ML models focussed on developer productivity and product growth - Propensity Model (suite of ML models built to identify the accounts that are most likely to benefit from the next tier of GitHub consumables), Repo Topics (Identify and tag the most relevant topics for a repository for improved developer experience) , Critical Path in Stochastic DAGs (detect bottlenecks in stochastic CI/CD pipelines). Established partnerships with product, growth and engineering teams and worked closely with other partner teams. Developed and communicated goals, strategies, project plans, timelines, and key performance metrics to reach goals. Filed a Patent for a proprietary algorithm - Critical Path in Stochastic DAGs. Mentored engineers and data scientists in the global team and contributed in defining a new hiring process for Data Scientists.

Development and successful deployment of multiple developer productivity models Failure Clustering for efficient troubleshooting, Association rules for missing files in PR, Intent Detection for PRs. Responsible for leading and delivering the V1.0 of Pipeline (CI/CD) Analytics for a Azure DevOps with 100K+ MAUs within the first month of release. Implementation of E2E MLOps practices for automated deployments, model tracking, data drift handling, testing and business validations. Develop strategies to help customers activate their analytics and create a more data-driven culture by providing in-context insights.

Product Manager for Azure SaaS service with responsibility to build E2E Test Analytics in Azure DevOps. The service handled 300+million test results in a single day of execution and catered to 1M+ MAUs. Product Manager Azure Mobile Engagement responsible for App Store Distribution service to publish MAM integrated apps on public stores. Program Manager in Strategic Enterprise Services leading the Self-Service app distribution to enable Enterprise Mobility at Microsoft.

Data Warehouse and BI applications providing metrics like Top Issues, Invoicing, Vendor Management, Agent Effectiveness for some of the most widely used products like Windows, Office, Zune etc. Handled 5TB per month of data inflow in the year 2008 when Relational databases were still predominant but had limited large size data handling capabilities. Built the acquisition pipeline for Enterprise Data Warehouse that served Microsoft Consumer and Enterprise Businesses.