Dec 2020 – Mar 2024 · 3 yrs 3 mos · Bengaluru, Karnataka, India

• Developing, Designing, Deploying, Migrating, Integrating, and Implementing end to end Deployment & Infrastructure strategies for the Machine Learning Platform.
• Actively participated in high-level team activities such as suggesting architecture improvements, recommending process improvements, and conducting tool evaluations.
• Developed concord pipelines to deploy Spark operators on Kubernetes clusters. • Design and Implemented GCS-SYNC solution in GKE for astronomer airflow (K8s
• executor).
• Installation of the GKE cluster via Terraform and configuring Workload Identity
• Management to pods to interact with other Google services.
• Databricks on Google Cloud Platform.
• Migration of Micro-services from self-managed Kubernetes cluster to GKE and AKS
• and exposing them via Istio.
• AKS and GKE Cluster Upgrade strategy for tier-0 MLOPS deployments.
• Design of Multi regional deployments for MLOPS.
• Implementing the effective cost attribution strategy to reduce GCP & Azure cloud
• cost for approximately 150 ML teams on the platform,
• Design and Implementation of Google TPU offering As a service in the Machine
• Learning Platform.
• In collaboration with the Google PSO, Design and Lead the implementation of
• Google Vertex-AI(Machine Learning service) offering As a service and integration
• with Walmart's Machine Learning Platform.
• Design and drive the implementation of multiple GPU flavors in Azure and GCP
• cloud, some of them are Nvidia A100, K80, T4, V100
• Developed Python Scripts to capture GPU utilization of different GPU flavors.
• Kubernetes cluster monitoring via Prometheus and Telegraf plugin.
• GPU enablement in self-managed Kubernetes.
• Serverless spark/Dataproc Batches in GCP
• Argocd
• MLFlow

Apr 2019 – Dec 2020 · 1 yr 8 mos · India

• Setup and Maintained 7+ of kubernetes cluster in GCP, Azure and On-premises.
• Automated kubernetes cluster setup in GCP and Azure using Terraform and Kubespray, reducing cluster uptime by more than 70%.
• Developed python script to prepull the docker images(size in GB) in the kubernetes clusters(300+ nodes), saving admin time by 80% and reducing pod start time by 95%.

Apr 2018 – Mar 2019 · 11 mos

• 1. Experience in setting up a multi-node Kubernetes cluster using kubernetes-incubator/kubespray.
• 2. Good understanding and experience in deploying different Kubernetes workflows.
• 3. Developed automation framework for deploying different kubernetes workflows using Kubernetes-Python SDK (kubernetes-client)
• 4. Developed Ansible-Playbooks for infrastructure automation.
• 5.Experience in Jenkins Master/Slave configuration with an end to end Jenkins Pipeline creation for CI/CD.
• 6. Build Automation using Jenkins, Scheduling Builds, Build Linking upstream and Down Stream and Parameterized Builds.
• 7. Experience in writing Docker files and building images from it.
• 8. Experience in Installation Cloudera Manager and CDH.
• 9. Monitoring infrastructure with various monitoring tools such as Splunk and Site24*7.

Feb 2016 – Apr 2018 · 2 yrs 2 mos

•  Involved in feature testing of the modules along with writing Test cases.
•  Developed Python scripts for automating Rest API.
•  Involved Manual testing of the web-based portal.
•  Developed a new framework using Python to validate the data stored in HBase using Phoenix Query
• Server (PQS).
•  Designing automation scripts in Robot framework using RIDE for GUI Automation and API Automation.
•  Preparation and execution of test scripts using JMeter for performance testing of REST API.
•  Involved in testing of REST API using Postman client.
•  Developed Python scripts for the testing end to end product flows, verifying the flow of the message
• through different components like Kafka, HBase and Ignite cache.
•  Developed a new framework using Python from scratch for executing REST API calls. This framework has a
• capability to auto-update test result after running test scripts and generates reports accordingly.
•  Developed Python scripts to monitor system resources usage and store the results to InfluxDB and
• Grafana tool is used to plot graphs and monitor the system resources using the results stored in InfluxDB.
•  Developed auto deployment scripts in a Shell for deploying micro-services successfully using Jenkins builds.
•  Designed and Implemented efficient Splunk dashboards using Splunk.
•  Developed and Managed CI/CD pipelines for applications.
•  Managed Build and Release activities for development teams