Senior AWS DevOps · SRE · Cloud Architect

NikhilSeemala

I build the delivery paths that get code to production and keep it running.

10+Years in cloud
35%Faster recovery
60%Faster provisioning
3Clouds in daily use
View experience Email me LinkedIn ↗
Nikhil Seemala
MTTR −35%
OIDC · no long-lived keys
How I think

Nothing changes by hand in production.

01 / Experience

Ten years of production AWS, role by role.

From ISP networks to streaming media and regulated data platforms. Each card shows the system behind the role. Open the record for the detail.

01

SiriusXM Canada NOW

Streaming media · AWS
AWS Cloud / DevOps Consultant

Owns the delivery path for AWS workloads, from the first pull request to Fargate and EKS services that stay healthy under load.

PR OIDC CDK Gate ECS rollback dev → staging → prod
Recovery time, approx.35% lower
ComputeFargate + EKS
CI to AWSKeyless
Oct 2023 – Present · Toronto, remote
Open the record →
  • Reusable GitHub Actions workflows that authenticate to AWS through OIDC, promoting CDK and CloudFormation stacks from dev to production with approval gates and rollback.
  • Tuned Fargate task sizing, health checks and scaling; traced flapping services through CloudWatch, Datadog and container logs.
  • Designed VPC layouts, endpoints, IAM and KMS standards; CloudFront and WAF in front of public traffic; on call for production incidents.
02

Airudi Inc.

Analytics platform · Hybrid cloud
Cloud Architect

Designed one analytics platform across AWS and Azure, built from modules that ship on their own.

AWS Azure network identity data · ETL Terraform modules
Release cycles, approx.50% shorter
CloudsAWS + Azure
DataPIPEDA
Apr – Oct 2023 · Montreal, remote
Open the record →
  • Built the platform in Terraform, split into modules so each workload could ship independently.
  • Set up GitHub Actions pipelines for infrastructure and application deployments.
  • Built KMS and Key Vault encryption plus role-based access for personal data, with Databricks ETL and secure REST APIs.
03

Global Intelligence

Cloud migration · Serverless
AWS Senior Solutions Architect

Moved workloads from on-premises to multi-region AWS under a strict rule: no manual change in production.

on-prem Region 1 Region 2 Terraform only
Provisioning, approx.60% faster
Business units12
Engineers mentored5
Jun 2021 – Apr 2023 · Fredericton, remote
Open the record →
  • Published reusable Terraform modules that cut provisioning time across 12 business units.
  • Built serverless APIs on Lambda, API Gateway and Cognito aligned to SOC 2 and NIST 800-53.
  • Ran architecture reviews and mentored five engineers on IaC patterns and operational practice.
04

Aphria Inc.

Data pipelines · AWS
AWS Cloud Engineer

Kept data moving into S3 and fixed the network problems that stopped it.

DMS Firehose S3 Databricks Glue network connectivity, fixed
ServicesGlue · DMS · Firehose
DestinationS3 + Databricks
Aug 2020 – Jun 2021 · Leamington, ON
Open the record →
  • Operated Glue jobs, Data Firehose streams and DMS replication into S3, with Databricks and PySpark downstream.
  • Diagnosed Glue connectivity failures across subnet routes, NAT, S3 endpoints, security groups and DNS.
  • Led enterprise migrations and Azure AD identity integration; automated provisioning and monitoring with Terraform and CloudWatch.
05

TekSavvy

National ISP · Networking
DevOps Engineer

Where the foundations came from: Linux, networking and automation at ISP scale.

Router Loadbalancer Linux Linux K8s VPN
Tenure3 yrs 8 mo
NetworkDNS · LB · VPN
Jan 2017 – Aug 2020 · Chatham, ON
Open the record →
  • Built Jenkins and GitHub Actions pipelines and Terraform automation for AWS.
  • Ran Kubernetes clusters and RHEL and Ubuntu hosts.
  • Managed DNS, load balancers, firewalls and VPN for a national ISP network.
02 / Capabilities

Where I operate.

The tools I use every week, grouped by the job they do.

Delivery

  • GitHub Actions
  • Jenkins
  • GitLab CI
  • ArgoCD
  • Azure DevOps
  • JFrog

Infrastructure as code

  • AWS CDK
  • CloudFormation
  • Terraform
  • Terragrunt
  • Bicep
  • Ansible

Containers

  • ECS Fargate
  • EKS · AKS
  • Docker
  • ECR
  • Helm
  • Auto Scaling

Observability

  • CloudWatch
  • Datadog
  • Prometheus
  • Grafana
  • ELK
  • OpenTelemetry

Network and security

  • VPC
  • NAT
  • Endpoints
  • Route 53
  • IAM
  • KMS
  • WAF
  • CloudFront

Data

  • Glue
  • Data Firehose
  • DMS
  • S3
  • Databricks
  • PySpark

Languages

  • Python
  • TypeScript
  • Bash
  • SQL

Clouds

  • AWS (primary)
  • Azure
  • Google Cloud
03 / How I work

How I keep production boring.

Five rules that show up in every platform I run.

R01

Nothing changes by hand in production.

Every environment comes from code. At Global Intelligence that was a hard rule across all environments, not a guideline.
R02

Rollback is rehearsed, not hoped for.

Promotion from dev to production carries approval gates and a tested way back, so a bad release is a short detour.
R03

Every alert points to a runbook.

An alert that does not say what to do next is noise. Tying Prometheus and Datadog alerts to runbooks cut recovery time by about 35%.
R04

No long-lived keys in CI.

Pipelines assume short-lived AWS roles through OIDC, with least-privilege IAM and KMS encryption underneath.
R05

Write it down for the next person.

Reusable modules, workflows and standards, so another team can adopt them without me in the room.
04 / Patterns

Patterns I reach for.

Four building blocks from the platforms I have run, drawn as simply as I can.

GitHub Actions AWS STS IAM role OIDC token, short-livedno stored keys
Pattern 1

Keyless CI to AWS

The pipeline proves who it is and receives a role that expires. Nothing secret sits in the repository.

dev staging prod rollback path, tested
Pattern 2

Promote with a way back

Each environment is a gate. A failed release returns to the last good state without a scramble.

VPC private subnet endpoint S3
Pattern 3

Private data paths

Workloads reach S3 and other services through VPC endpoints, so traffic stays off the public internet.

alert runbook fix Prometheus · Datadogabout 35% faster recovery
Pattern 4

Alerts that say what to do

Every alert links to the steps that resolve it, so the person on call starts with an answer.

Scroll sideways →
05 / Credentials

Certified at the professional level.

AWS Certified Solutions Architect Professional badge
AWS Certified Solutions Architect – ProfessionalAmazon Web Services
Google Cloud Professional Cloud Architect badge
Professional Cloud ArchitectGoogle Cloud
EducationB.Tech, Computer Science and EngineeringJawaharlal Nehru Technological University
2010 – 2014
06 / Contact

Have a platform to ship and keep running?

Contract or full-time, remote. I work from Canada and take US roles on a TN. Send a note and I will reply with my resume.

Message on LinkedIn
I read every message and reply with my resume.