I'm Emre Aydinsoy, an independent DevOps and cloud consultant. For 7+ years I've run production AWS for products with millions of users. Now I do the same for your team: architecture, automation, cost and AI. You work directly with me, with no account managers, no juniors and no handoffs.
Services
Fixed-scope projects or ongoing support. I work hands-on in your accounts and repos, and everything I build ships as code your team owns.
Generative AI wired into your operations, inside your own AWS account. Your data stays in your VPC, and every agent is versioned and deployed like the rest of your infrastructure.
On-prem to AWS, account consolidations and re-architectures without drama.
EKS and ECS platforms that developers actually enjoy deploying to.
No more click-ops. Every environment reproducible and reviewed.
FinOps that pays for itself, usually within the first month.
Audit-ready AWS without slowing the team down.
Know before your customers do, and sleep through the night.
Retire servers you shouldn't be patching anymore.
The boring, repeatable work, done by scripts instead of people.
Selected work
A few recent engagements. Details are shared with permission or kept anonymous.
Moved 9 on-prem servers to ECS across 3 availability zones with host-based ALB routing. Modernized the data layer to Aurora MySQL and ElastiCache, retired CentOS 6 and PHP 5, and connected the office over site-to-site VPN.
Designed a fully private three-tier VPC with dual-target flow logging and live-tested alarms. Containerized 8 production services with GitHub Actions to ECR, health-checked releases and versioned rollbacks.
Merged three production AWS accounts (60+ EC2, EKS and ECS clusters, Aurora fleets) into one multi-VPC account with WireGuard and phased DNS cutovers, while cutting monthly spend by 30–35%.
How we work
Every engagement starts with a free 30-minute call. You get a written plan before any work begins.
The person on the call is the person writing your Terraform. Nothing gets lost between sales, delivery and support.
No sprint-planning overhead. Decisions in Slack the same day, changes shipped in your repos as pull requests.
Pipelines, scripts and AI agents do the repetitive work, so one person can cover what usually takes a team.
Review your accounts, pipelines and bills. Find the quick wins and the real risks.
A short written roadmap with priorities, effort and expected savings.
Hands-on delivery in your repos, as code, with weekly demos.
Docs, runbooks and a walkthrough so your team owns it.
Toolbox
Let's talk
Tell me what you're working on. I reply within one business day, in English or Turkish.