Terraform Drift Detection
Python tool that polls Terraform state and compares it against live Azure resources, alerting on configuration drift before it becomes an incident. Cut drift incidents 70% and prevented two critical outages.
Available for work
Infrastructure
as code.
Deploys as routine.
I'm Tejas Mehta — a DevOps engineer with 3+ years on Azure, building CI/CD pipelines, Terraform modules, and AKS platforms that teams can actually trust.
45%
Faster release cycles
99%
AKS deploy success
2 hrs
To provision an env
70%
Fewer drift incidents
About
Automate the boring.
Ship with confidence.
I'm a DevOps engineer at Accenture with 3+ years designing, automating, and operating cloud infrastructure on Azure. My work sits where CI/CD, Kubernetes, and infrastructure-as-code meet — and my bias is always toward the boring outcome: predictable deploys, reproducible environments, and a quiet on-call rotation.
I hold a degree in Electronics & Communication Engineering and am Microsoft certified in Azure (AZ-104, AZ-900).
Azure at scale
AKS platforms spanning 60+ services, lifted from 85% to 99% deployment success.
Everything codified
Terraform modules that took environment provisioning from 4 days down to 2 hours.
Security in the pipeline
OPA/Rego, Gatekeeper, SonarQube and Snyk gates preventing 95% of policy violations at PR time.
Cost kept honest
InfraCost checks and right-sizing that cut £1,200/month of waste and 40% of agent spend.
Toolbox
One toolchain.
End to end.
Cloud (Azure)
IaC & Containers
CI/CD
Observability
DevSecOps
Languages & OS
Process
Four steps.
No surprises.
Design
Map requirements to a reproducible architecture. Modules, environments, and guardrails get defined before a single resource is provisioned.
Codify
Everything as code — Terraform for infra, Helm for workloads, OPA/Rego for policy. Reviewed in PRs, versioned, and cost-estimated up front.
Automate
End-to-end CI/CD with security and quality gates. Deployment gating, drift detection, and self-service pipelines teams actually trust.
Observe
Close the loop with Application Insights and the ELK stack. Autoscaling, alerting, and continuous tuning to keep on-call quiet.
Portfolio
Things I've built.
And why they mattered.
Policy-as-Code Platform
OPA/Rego policy library with a Gatekeeper admission controller enforcing enterprise compliance on AKS. Prevents 95% of policy violations at PR time and generates auditable compliance reports.
This Portfolio
Static Astro site on a private S3 bucket behind CloudFront. Infrastructure fully described in Terraform; zero long-lived AWS keys thanks to GitHub Actions OIDC. Initial payload under 100 KB.
Writing
Notes from the
pipeline.
Building a Terraform Drift Detector in Python
Configuration drift is silent until it causes an outage. Here's a lightweight tool that catches it before your next apply does.
Read post →Cutting Release Cycle Time by 45% with Azure DevOps
How reusable pipeline templates, deployment gating, and parallelised stages took our microservice releases from slow and manual to fast and boring.
Read post →