Cloud infrastructure · Automation · Reliability

Hiren Timbadiya

DevOps Engineer at ArmorCode

I build cloud platforms that ship faster, cost less and stay up.

I automate AWS infrastructure, CI/CD, observability and disaster recovery for SaaS platforms.

Get in touch Download resume

Engineering impact 1h → 0 Daily release call eliminated

Daily tools

AWS
Kubernetes
Terraform
Docker
Helm
Jenkins
GitHub Actions
Prometheus
Grafana
ELK

Every tower is a system

Systems I built and run. For each one: the problem, the build and what it changed.

Release automation · AWS

In production

1h → 0

daily release call, deleted

Remote release runner

Problem: Every customer release needed a one-hour call to run Jenkins jobs by hand.
What I built: Lambda and API Gateway take a JSON list of Jenkins jobs and store it in DynamoDB. The customer-side Jenkins polls the table, runs the jobs in order and reports status back.
Production impact: The daily release call is gone. Releases run on their own and report back.

Lambda
API Gateway
DynamoDB
Jenkins

Release automation · AWS Remote release runner 1h → 0 daily release call, deleted

Problem
Every customer release needed a one-hour call to run Jenkins jobs by hand.

What I built
Lambda and API Gateway take a JSON list of Jenkins jobs and store it in DynamoDB. The customer-side Jenkins polls the table, runs the jobs in order and reports status back.

Production impact
The daily release call is gone. Releases run on their own and report back.
- Lambda
- API Gateway
- DynamoDB
- Jenkins
Disaster recovery · AWS Cross-region backup & DR 3 data stores backed up across regions and accounts

Problem
Stateful data in RDS/Aurora, DocumentDB and OpenSearch needed copies outside one region and one account.

What I built
EventBridge and Lambda copy backups across regions and accounts. KMS encrypts every copy, and automated health checks watch the pipeline.

Production impact
All three data stores have encrypted copies in another region and another account.
- EventBridge
- Lambda
- KMS
- RDS / Aurora
- DocumentDB
- OpenSearch
Cost optimisation · AWS Graviton migration $960 saved every month (−11%)

Problem
The x86 compute bill was $8,600 a month.

What I built
Moved the workloads from x86 to ARM Graviton instances.

Production impact
The bill went to $7,640 a month with zero performance regression.
- AWS Graviton
- EC2
- ECS
- Docker
CI/CD · Jenkins on ECS On-demand CI agents +80% CI/CD throughput

Problem
Jenkins needed more build capacity without paying for idle agents.

What I built
Moved Jenkins builds to on-demand ECS/EC2 agents.

Production impact
About 80% more CI/CD throughput, and about $224 a month less on agents.
- Jenkins
- ECS
- EC2
- Terraform
Autoscaling · ECS + SQS Queue-depth autoscaling $300 idle compute cut every month

Problem
The AI-correlation workers kept running when there was no work.

What I built
ECS step scaling driven by SQS queue depth: workers scale out when work arrives and back down when it stops.

Production impact
About $300 a month of idle compute removed.
- ECS
- SQS
- CloudWatch
- Terraform
IaC quality · GitHub Actions Terraform PR guard Every PR checked before merge

Problem
Terraform misconfigurations and policy drift could reach the main branch.

What I built
GitHub Actions pipelines run AI-powered review and tflint on every Terraform pull request.

Production impact
Misconfigurations and policy drift get caught before merge.
- GitHub Actions
- tflint
- Terraform
Observability · ELK + Grafana Self-hosted observability −60% incident response time

Problem
Logs lived in a licensed tool (Logz.io) with per-host costs.

What I built
Self-hosted ELK plus Prometheus and Grafana: Logstash pipelines, ILM hot/warm/cold tiers and S3 snapshots.

Production impact
Incident response got about 60% faster and the per-host licences went away.
- Elasticsearch
- Logstash
- Kibana
- Prometheus
- Grafana
Cloud security · IAM IAM clean-up −40% overprivileged IAM roles

Problem
Too many IAM roles had more access than they needed, and API keys were hardcoded.

What I built
Quarterly security audits, tighter policies and security groups, and keys moved into Google Secret Manager.

Production impact
40% of overprivileged roles removed across dev and prod, and a credential-exposure gap closed.
- IAM
- Security groups
- Secret Manager

The night line

Newest stop first: the job, the numbers, the work.

May 2024 – now
ArmorCode Inc.

Gurugram
1. DevOps Engineer Jul 2025 – now
2. Associate DevOps Engineer Nov 2024 – Jun 2025
3. DevOps Intern May 2024 – Nov 2024
- 1h → 0 daily release call
- $960/mo saved with Graviton
- +80% CI/CD throughput
- $300/mo idle compute cut
- Cloud, CI/CD and observability for a SaaS security platform.
- Built the remote release runner and cross-region, cross-account backup and DR.
- Run the AWS Load Balancer Controller on EKS, with ALB ingress routing to 3 internal services.
- Maintain Helm charts with per-environment values for qa, staging and prod.
Nov 2023 – Apr 2024
Software Engineer Intern, DevOps

LaNet Team Software Solution · Surat
- −40% overprivileged IAM roles
- Quarterly security audits that hardened IAM policies and security groups in dev and prod.
- Prometheus and Grafana dashboards for service health.
- Moved hardcoded API keys into Google Secret Manager.
Mar 2023 – Oct 2023
Open Source Contributor & Community Member

Layer5 · Remote · Austin, Texas
- 12 PRs merged into Layer5 and Meshery
- Active contributor and community member across Layer5 open-source projects.
- Spent time reviewing other contributors’ pull requests.
- Helped keep the community a place where people share what they know.
2020 – 2024
B.E. Information Technology

L.D. College of Engineering · Ahmedabad
- CGPA 8.66 / 10.

Projects

Lights on after hours

Products I'm building on my own time, and infrastructure code anyone can use.

Featured · Desktop app · Rust + Tauri

Private

OpsBastion

Run your Linux servers from one desktop app.

Looking after servers means juggling SSH tabs, scp and htop, and AI tools have no safe way to look in.

One window: terminal, files, Docker, services, cron, logs, metrics, alerts and runbooks.
A read-only MCP server with secrets masked, so AI tools can look but never change.
Private EC2 with no open port or key file: SSH over Session Manager with one-time in-memory keys.

faster folder loads (2+ min → 7 s): 17×
to open a 130 MB log: <100 ms
checks on 4 Linux distros: 900+

Rust
Tauri
React
TypeScript
SQLite
AWS SDK
MCP

Private repo

File system · Rust

Building

AnywhereFS

Cloud drives that mount like a local disk.

Problem Cloud drive files should open in any app, like files on a local disk.
- A Rust engine serves Google Drive over a local NFS mount.
- A Tauri desktop app sits on top.
- Rust
- NFS
- Tauri
- Google Drive API
Code not public yet

Open infrastructure code

hirentimbadiya/aws-network-firewall-terraform

Terraform · Network security

Open source

AWS Network Firewall

Egress filtering you can roll out safely.

Problem Locking down outbound traffic can break production on day one.
- Multi-AZ firewall subnets with NAT-based egress routing.
- A domain allowlist on HTTP_HOST and TLS_SNI.
- An enable_strict_mode switch moves from monitor to drop.
- Terraform
- Network Firewall
- VPC
- CloudWatch
GitHub : AWS Network Firewall
hirentimbadiya/ecs-sqs-autoscaling-terraform

Terraform · Autoscaling

Open source

ECS + SQS autoscaling

Scale workers from queue depth, down to zero.

Problem AWS has two ways to scale ECS from SQS, and most guides show only one.
- Step scaling and target tracking side by side, in Terraform.
- Scale-from-zero with CloudWatch metric math (visible + in-flight).
- Terraform
- ECS Fargate
- SQS
- CloudWatch
GitHub : ECS + SQS autoscaling
hirentimbadiya/elk-stack-aws-terraform

Terraform · Logging

Open source

ELK on AWS

Production logging, fully in Terraform.

Problem A self-hosted log stack needs a lot more than three EC2 boxes.
- Private subnets, with Nginx and TLS in front of Kibana.
- Four Logstash pipelines with GeoIP, ILM tiers and S3 snapshots.
- Terraform
- Elasticsearch
- Logstash
- Kibana
GitHub : ELK on AWS

Lights in other cities

My code, merged into other people's projects.

Pull requests merged into other projects: 17
Open-source projects contributed to: 6
Stars on my repositories: 35
Followers on GitHub: 58
Public repositories: 37

Merged into layer5io keptn meshery grafana open-feature

A few pull requests

Counted from the GitHub API on 2026-11-01. See it all on GitHub.

Notes from the night shift

What I learn running production, written down.

Opening soon

The first notes from the night shift are being written: Graviton, DR drills and what broke along the way. Follow @heyhiru to hear when they land, or grab the RSS feed.

Hiren Timbadiya in a dark suit and glasses, smiling, in an office with the city behind the glass. — On call · Gurugram, India

Who's on call

I'm a DevOps engineer with 2+ years on AWS: cloud automation, CI/CD and infrastructure as code. I like systems that are reliable, cheap to run and easy to operate.

At ArmorCode I run the platform behind a SaaS security product, from provisioning to observability to what happens when something fails at 2 AM. I also moved our logs off a licensed tool to a self-hosted stack and cut incident response by about 60%.

I studied Information Technology at L.D. College of Engineering, Ahmedabad (CGPA 8.66).

Cloud & IaC: AWS · GCP · Terraform · CloudFormation · Ansible · Packer
Containers: Docker · Kubernetes · Helm · ECS · EKS · ArgoCD
CI/CD: Jenkins · GitHub Actions · GitLab CI · GitOps
Observability: Prometheus · Grafana · ELK · Loki · New Relic · CloudWatch
Security: SonarQube · Semgrep · Trivy · Prowler · KMS · IAM
Code: Python · Bash · Go · Java · Linux

Off shift: Cricket, The stock market, AI and LLMs, Travel, Music, Gaming, Volunteering.

05:42 IST · shift over

Got a system that keeps you up at night?

Full-time roles, freelance infra work, or a second opinion on your AWS bill. Tell me what's breaking.

Available for opportunities Open to DevOps, SRE and platform engineering roles, and freelance infra work.

Email me Download resume

hiren@hirentimbadiya.com

Hiren Timbadiya

Every tower is a system

Remote release runner

Cross-region backup & DR

Graviton migration

On-demand CI agents

Queue-depth autoscaling

Terraform PR guard

Self-hosted observability

IAM clean-up

The night line

ArmorCode Inc.

Software Engineer Intern, DevOps

Open Source Contributor & Community Member

B.E. Information Technology

Lights on after hours

OpsBastion

AnywhereFS

Open infrastructure code

AWS Network Firewall

ECS + SQS autoscaling

ELK on AWS

Lights in other cities

A few pull requests

Notes from the night shift

Who's on call

Got a system that keeps you up at night?