Open to work · Chandigarh, India

Kartik

DevOps Engineer · Multi-Cloud · Kubernetes

$ terraform apply▋

View Work Download Resume GitHub

Faster provisioning

Terraform + Bicep IaC

Faster incident response

ELK + Python automation

Lower cloud spend

Right-sizing + S3 lifecycle

Architecture I Operate

Multi-cloud delivery: a single CI/CD path ships 20+ Helm-managed microservices to AWS EKS and Azure AKS, with centralized ELK observability.

Projects

FuseGuard — Circuit Breaker for AI Agents

Problem: AI agents can run away and burn thousands in API spend before anyone notices; dashboards show spend but don't stop it.

Approach: Built a drop-in proxy that enforces token/$ budgets at runtime on Cloudflare Workers + Durable Objects — reserves worst-case cost pre-flight, reconciles after, kills runaway loops. Open-core: MIT proxy + hosted Next.js dashboard, Supabase (RLS), Lemon Squeezy billing, full CI/CD.

Shipped end-to-end and deployed to production; 168 tests, two Opus-grade security audits, fail-closed budget enforcement with no race conditions.

Cloudflare Workers
Durable Objects
Next.js
Supabase
TypeScript

Code →

Kubernetes Multi-Cluster Management with Helm

Problem: 20+ microservices deployed inconsistently across EKS and AKS.

Approach: Standardized Helm charts with per-environment value overrides and automated release workflows across dev, staging, and production.

90% less manual deploy effort; zero-downtime rolling deployments.

AWS EKS
Azure AKS
Helm
Kubernetes

Multi-Cloud Infrastructure-as-Code Pipeline

Problem: Slow, drift-prone environment provisioning across two clouds.

Approach: Reusable Terraform (AWS) + Bicep (Azure) modules for networking, compute, and Kubernetes resources, version-controlled for org-wide adoption.

One-click provisioning; 95% faster setup; enforced standard patterns.

Terraform
Bicep
AWS
Azure AKS

ELK Log Analysis & Monitoring Automation

Problem: Manual debugging slowed incident response across distributed services.

Approach: Python tooling for centralized log analysis on the ELK stack, with structured logging, proactive alerting, and centralized monitoring.

40% faster incident response; reduced manual debugging effort.

Python
Elasticsearch
Logstash
Kibana

Tech Arsenal

Cloud

AWS (EC2, RDS, S3, IAM, EKS, ECR)
Azure (AKS, ACR, Monitor)

Containers & Orchestration

Docker
Kubernetes
Helm

Infrastructure as Code

Terraform (AWS)
Bicep (Azure)

CI/CD

GitHub Actions
Azure DevOps Pipelines

Observability

ELK Stack
Amazon CloudWatch
Azure Monitor

Scripting

Python
Bash

Databases

MySQL
PostgreSQL
MongoDB
DynamoDB

Experience

DevOps Engineer · Quark Software Inc.

Jul 2025 — Present

Architected a Kubernetes microservices platform of 20+ services across AWS EKS and Azure AKS using Helm, ensuring high availability and zero-downtime rolling deployments.
Engineered multi-cloud Infrastructure as Code with Terraform and Bicep, cutting environment provisioning time ~80%.
Developed Python automation for centralized ELK log analysis, cutting incident response time ~40%.
Optimized AWS resource configuration across EC2, RDS, S3, and IAM, reducing monthly cloud spend by 20%.

DevOps Trainee · Quark Software Inc.

Jan 2025 — Jun 2025

Containerized application services with Docker and deployed to managed Kubernetes (EKS, AKS).
Built reusable Terraform + Bicep modules for VPCs, compute, and clusters, reducing setup time 90%.
Implemented end-to-end CI/CD pipelines with GitHub Actions and Azure DevOps.

About

DevOps Engineer with hands-on experience designing and operating cloud-native infrastructure across AWS and Azure. I build reproducible IaC, reliable Kubernetes platforms, and observable systems that ship faster and cost less.

▹ Research paper presented at ICSPN, Dubai 2023
▹ Winner — Innovative Product Design, IPD-EXPO 2022
▹ B.E. Computer Science — GPA 8.73 / 10.0