Redmond, Washington, United States
March'2024-present: Working in the AI Hypercomputer team which is bringing the latest GPUs and TPUs to the GCP customers. Support for GKE in cluster toolkit: Led a team of 6 engineers (and many cross functional leads) across time zones (USA, Poland and India) to plan and launch support for GKE in cluster toolkit (with support of A3 Mega and A3 Ultra GPUs) and worked with external customers to onboard them to the platform: https://cloud.google.com/ai-hypercomputer/docs/create/gke-ai-hypercompute Aug' 2022-March'2024: Founding member and TLM of Unified Maintenance Management (UMM) and got the product to Private preview : https://cloud.google.com/unified-maintenance/docs/overview
Worked on ObjectStore at Dropbox which is a managed object store on top of internal store and AWS S3. More information here https://dropbox.tech/infrastructure/abstracting-cloud-storage-backends-with-object-store
Working in GCP Memorystore since March 2018, where I joined the product at an early stage (alpha. This has helped me to understand the work that needs to be done to productionize a Managed service, which includes (but is not limited to) monitoring, observability for both Customers, Engineers/Ops team and Customer Support Engineers. Have scaled up the service multiple times since the inception. Have been working in Google Cloud ever since June 2016, where I initially worked at AppEngine Flexible from June 2016 to March 2018. Owned the system which releases OS/internal system updates (gradually) to all the customer code without disrupting running applications. Also worked with a team on the Control Plane of App Engine Flexible to modernize the stack to run on latest technologies inside Google.
Working in the Azure Data Catalog (https://azure.microsoft.com/en-us/services/data-catalog/) team which deals with discovery of Metadata. My role is to deliver the Indexing and Search of the product, using Elasticsearch cluster running in Microsoft Azure. My team's job is to define how the data is indexed in the cluster in order to power the various search scenarios. Have worked on this product offering from inception to Public Preview, which has given me the opportunity to understand the design decisions necessary to take a product from its inception to market.
Designing and developing ecosystem for MINTS3 project of the Department of Epidemiology, University of Minnesota. Maintaining and enhancing the existing modules to manage users in Django framework. The other important parts of the framework are automated reminders to each participant when the survey's are active, keeping track of progress and payments. Application is entirely built on django framework of python and uses MySQL as its backend. Also I have limited integration responsibility with limeSurvey (in PHP) which administers the surveys. Main Technologies Used Django (python) MySQL
Relevant Coursework Analysis of Advance Algorithms Highly Available and Scalable Distributed Systems. Principles of Programming Languages. Principles of Database Management. Artificial Intelligence