Baskar Jayakumar

• Manage 25+ on-premise & 100+ EKS kubernetes clusters that comprise more than 3000+ nodes handling 10K jobs & petabytes of data per day.
• Manage all Spark infrastructure that runs on top of Kubernetes.
• Build & maintain deployable Spark versions from the open-source community for internal customers.
• Patch open-source Spark with security vulnerability fixes.
• Implemented HPA (Horizontal Pod Scaling) to resolve persistent issues related to resource overhead.
• Implemented automated deployment of kubernetes to on-premise clusters using AWS code deploy.
• Implemented cost-saving measures by adopting optimized hardware & software improvisations in EKS & AWS.
• Conducted root-cause analysis after major incidents to identify areas for process improvement.
• Collaborate with stakeholders to optimize infrastructure performance.

• Created full stack Hadoop clusters, EMR clusters, Qubole clusters & Kubernetes cluster.
• Implemented Ranger on cloud Data Lakes to mask PII & PCI data.
• Upgraded Hadoop Clusters from Cloudera CDP to Hortonworks HDP.
• Implemented R, RHive, R Studio in Hadoop Clusters.
• Implemented High Availability Servers for Namenode, Resource Manager, Hive.

Open Source Project contributions :

Jetstream - Data Replication tool to move data from on premise to Cloud.
Data Highway - Data Streaming platform to stream data from on-premise to Cloud.
Apiary - Tool to provision Data Lakes in Cloud.

• Designed & Provisioned HDP 2.X Hadoop Clusters via Ambari automation.
• Administrator of 8 Hadoop clusters with 500+ nodes & 2.5+ Peta Bytes data.
• Upgraded Hadoop version from Cloudera (CDP) to Hortonworks (HDP)
• Implemented High Availability set-up on Namenode, Resource Manager, Hive.
• Encrypted PII/PCI data using IBM Gaurdium tool.
• Automated Data replication on critical data to Disaster Recovery Cluster.
• Led the offshore team.

• Managed 4000+ production POS servers of Home Depot stores.

• Built new servers for production by applying relevant packages & softwares.