I keep production Kubernetes running when nodes die.
Now at Copado: autoscaling, cost and observability across multi-region production infrastructure. Five years on AWS and GCP, before that at BLG and Treebo.
cluster · simulated
- nodes ready
- 9 / 9
- pods running
- 24
- requests served
- 12,480
- failed requests
- 0
- last recovery
- —
Proof
- 40%cut off the monthly cloud billKubernetes rightsizing, HPA, infrastructure audits
- 2 TBmigrated live, with one planned cutoverAWS DMS driven by Python automation, with structure and row-count verification
- >50%smaller alert payloadsCustom Slack alert templates, first in Grafana and later in Coralogix
- 80sfrom pending pod to a ready nodeKarpenter in place of fixed node groups, tuned against observed behaviour
- cluster operations without leaving SlackSlack bot with verified requests, role-scoped commands and an audit channel
- a drop-in mail CLI on SESSame stdin interface as the tool it replaced, AWS SES underneath
Experience
Infrastructure Engineer, Cloud Operations
Copado · Jaipur, India
Own the platform across development, QA and multi-region production environments. Took Karpenter from a proof of concept through to every production environment, retiring the Cluster Autoscaler node groups it superseded. Stood up production environments end to end — Terraform, networking, secrets, registries and CI/CD — and wrote up the process so the team can repeat it.
Cloud Architect (contract)
BLG Technologies · India
Designed AWS architectures for clients in regulated sectors under a security-first model — account structure, VPC design and service selection. Set IAM governance with environment isolation and role-based access boundaries, and defined the monitoring and alerting baseline that gated go-live.
DevOps Engineer
Treebo Hospitality Ventures · Bengaluru, India
Co-owned the production EKS platform — dozens of microservices with RDS and Redis behind them, across staging and production. Owned stability, observability, cost and ISO audit readiness, and carried on-call through releases and incidents. Built Jarvis, a Slack bot giving developers and QA role-scoped access to the cluster without leaving Slack.
Associate DevOps / SRE Engineer
Treebo Hospitality Ventures · Bengaluru, India
Built and operated the AWS EKS platform — VPC architecture, subnetting, routing and multi-AZ deployment for staging and production. Drove cost work across the estate: RDS resizing and relocation, EC2 right-sizing, IOPS tuning, S3 lifecycle rules, and scheduled scale-down of non-production overnight and at weekends, with a self-service path for developers to bring environments back up on demand.
SRE intern
Treebo Hospitality Ventures · Bengaluru, India
Recognition
Star of the Month
Introducing Karpenter autoscaling, and production root-cause analysis.
CEO recognition
Cited for anticipating needs and moving work forward without being asked.
Employee of the Month, twice
Infrastructure alerting improvements, and cloud cost optimisation.
Stack
- Kubernetes, EKS and Docker
- Production clusters continuously since 2021, across development, QA and multi-region production.
- Karpenter
- Took it from a proof of concept through to every production environment, retiring the Cluster Autoscaler node groups it superseded.
- Terraform
- Effectively the whole estate is managed as code — networking, clusters, IAM, secrets, registries and CI/CD. Observability alerting is under Terraform now too.
- Helm
- Applications and platform components ship as charts. Day to day that means authoring, releasing and upgrading them in production.
- AWS
- EKS, EC2, RDS, S3, Lambda, ECR, Route 53, WAF, Secrets Manager, SES, CloudWatch and DMS across multi-account estates.
- GCP
- Production workloads alongside AWS.
- OpenTelemetry and Coralogix
- Consolidated log routing onto OpenTelemetry and tuned labelling to keep ingestion efficient. Built custom Coralogix-to-Slack alert templates: a severity title, one line on what the alert is for, and only the fields needed to act.
- Prometheus, Grafana and ELK
- Metrics, dashboards and alerting for the services I run — built to improve MTTR, not to fill a wall.
- AWS SSO and IAM
- Multi-account governance with enforced production isolation, least-privilege roles, application-scoped access and permission sets.
- GitHub Actions and Jenkins
- Build and deploy pipelines for application and infrastructure repositories, including maintaining legacy Jenkins jobs.
- Python
- Operational tooling — a Slack bot for safe cluster access, and release-verification automation now used by the team.
- Networking
- VPC design, subnetting and routing, NAT and internet gateways, load balancing and WAF, multi-AZ across staging and production.
The commands I reach for are public too: my cloud runbook, over 250 of them, searchable.
Certifications and education
AI fluency and Claude development
- AI Fluency: Framework & Foundations
- Claude 101
- Claude Platform 101
- Claude Code 101
- Claude Code in Action
- Introduction to Agent Skills
- Introduction to Claude Cowork
Copado AI
Bachelor of Engineering, Computer Science
Work with me
Yash Verma
Jaipur, India
LinkedIn is the fastest way to reach me. If you would rather email, the address copies to your clipboard with one click.