Varshini A S

DevOps Engineer with 3+ years of hands-on experience building, securing, and scaling AWS infrastructure. Designed VPC networking architecture from scratch (public/private subnets, VPC peering, VPC endpoints), built a Packer/Terraform immutable infrastructure pipeline, and automated AMI creation with a standalone Python (FastAPI) service and Bash tooling. Experienced with GitHub Actions and Jenkins CI/CD, Kubernetes (EKS operations and self-managed kubeadm clusters), and observability (Grafana, Loki, Prometheus, CloudWatch).

Promoted from Junior DevOps Engineer to DevOps Engineer after leading the VPC and security hardening initiative, and now mentor team members on AWS and DevOps practices.

DevOps Engineer at StorTrack (formerly Aggregate Intelligence, Inc.) (2026-07 – Present)

Promoted from Junior DevOps Engineer to DevOps Engineer after leading the VPC redesign and security hardening initiative below; now one of 3 engineers on the infrastructure team and mentor colleagues, including a senior team member new to AWS/DevOps.

• Designed AWS VPC architecture from scratch with public/private subnet isolation and configured VPC endpoints (CloudWatch, S3, Lambda, SNS) so private resources could reach those services without internet exposure.
• Migrated application servers, database servers, and S3-dependent workloads into the redesigned VPC's private subnets.
• Configured VPC peering between two AWS accounts to enable secure cross-account connectivity.
• Configured a VPN on an EC2 instance for secure remote access to internal resources.
• Replaced roughly 50 long-lived IAM access keys with IAM roles across the environment, and removed SSH access from 100+ servers in favor of AWS Systems Manager, closing a credential-exposure risk and an open port 22 exposure identified during a prior server migration.
• Configured and secured Apache Kafka and Zookeeper clusters with SASL/SCRAM authentication and ACL-based authorization for user-level access control.
• Built a self-updating AMI pipeline with Packer — pre-baking Docker pull commands, the CloudWatch agent, NGINX, and Grafana Alloy into each image — feeding a Terraform-managed infrastructure stack (launch template, Auto Scaling Group, Load Balancer, Target Group, Security Groups) across staging and production.
• Built a separate Python (FastAPI) service to automate AMI creation, independent of the Terraform workflow.
• Wrote Bash tooling to automate parts of the Terraform workflow, reducing manual steps in applying infrastructure changes.
• Initiated structured Ansible roles (replacing ad hoc playbooks) and containerized node deployments that previously ran directly on bare servers.
• Converted GitHub Actions CI/CD pipelines into parameterized, reusable workflows called from a central repository, removing duplicated pipeline code across services.
• Operated Amazon EKS clusters, troubleshooting pod out-of-memory failures and storage issues caused by insufficient resource allocation, and configuring CloudWatch Container Insights for cluster-level monitoring.
• Diagnosed high database server load during a new product release by narrowing the issue down to a single API endpoint driving excessive load, then identifying idle connections as a root cause; introduced PgBouncer for connection pooling and tuned PostgreSQL logging and configuration to surface similar issues earlier, while partnering with the database team on query optimization.
• Diagnosed degraded read/write performance on a separate PostgreSQL server caused by dead tuple buildup from continuous truncate-and-insert operations; resolved it by tuning autovacuum worker settings, restoring normal query performance.
• Built centralized observability with Grafana Alloy shipping logs to Loki, Grafana dashboards for visualization, and a PostgreSQL exporter for database-specific alerting; set up Grafana alerts for database CPU/memory usage and CloudWatch monitoring for infrastructure health.
• Extended observability into Kubernetes by configuring Grafana Alloy for EKS log collection and enabling Container Insights on a separate cluster.
• Migrated 250+ company domains from GoDaddy to AWS Route 53 with zero downtime.
• Migrated servers and S3 buckets — some exceeding 2TB — across 4 AWS accounts using AWS DataSync.
• Administered PostgreSQL databases — managing user access, tuning configuration, and handling backup scheduling, restore, and dump processes.
• Continue right-sizing infrastructure based on CPU/memory usage and using AWS Cost Explorer to identify and eliminate unused billed resources.

Junior DevOps Engineer at StorTrack (formerly Aggregate Intelligence, Inc.) (2024-06 – 2026-06)

DevOps Intern at StorTrack (formerly Aggregate Intelligence, Inc.) (2023-09 – 2024-05)

• Deployed a centralized Prometheus setup to monitor CPU and memory usage on database servers (MongoDB, PostgreSQL), with Grafana alerts emailed when utilization exceeded 95% for 10+ minutes.
• Managed Ansible automation across 100+ servers.
• Moved Jenkins from single-node to multi-node.
• Built foundational experience with Docker, AWS, and GitHub Actions, carried forward into the Junior DevOps Engineer role.

Bachelor of Engineering in Computer Science Engineering – Knowledge Institute of Technology (KIOT) (2020 – 2024)