Mohammad-Amine Banaei - Go & Platform Engineer
I write Go services, automate Kubernetes platforms, and hand over clear documentation and training so your team doesn't depend on me. 5 years of delivering systems that stay up and teams that stay autonomous.
How I Improve Your Systems
Performance
I replace slow, serial code paths with concurrent Go pipelines that slash processing time and cost.
Stability
I turn risky manual deploys into a frictionless GitOps loop where every change is tested, reviewed, and rolled out automatically.
Autonomy
I never leave a project without a runbook, a workshop, and a team that can operate the system as well as I can.
Ways I Can Help
Go Backend Development
For teams that need high‑performance APIs or want to migrate hot paths from Node/PHP to Go.
- Clean Architecture, gRPC, or REST.
- CQRS and Event‑Driven patterns when needed.
- CI/CD integrated from day one, with automated tests and deployment to a staging cluster.
Kubernetes & Infrastructure Automation
For startups ready to move from manual deploys to a professional platform.
- HA Kubernetes clusters (cloud or bare‑metal) with Argo CD GitOps.
- Observability stack (Prometheus, Grafana, Loki, Jaeger, OpenTelemetry).
- Infrastructure as Code with Terraform, repeatable and auditable.
Training & Knowledge Transfer
For engineering teams that want to own their stack.
- Tailored courses in Go, Kubernetes, and CI/CD (200+ engineers trained).
- Operational runbooks and one‑page "how to restart X" guides.
- Recorded walkthroughs and live Q&A sessions - you become independent fast.
Real Work, Real Numbers
1M+ CSV Rows in <3 Minutes
The Challenge
Critical business data was stuck behind slow pipelines; legacy exports took >15 minutes and required manual reconciliation.
The Solution
Designed a highly concurrent Go ingestion pipeline on AWS (Lambda + ECS Fargate + SQS). Parallelised MongoDB queries and implemented deduplication logic.
The Result
1,000,000+ rows processed in 3 minutes, report generation reduced to <2 minutes, zero duplicates.
Stack
Go, AWS Lambda, ECS Fargate, MongoDB Aggregations, SQS
Deployment Failure Rate 40% → <5%
The Challenge
Manual, chaotic release process with frequent production breakage; engineers spent hours fixing deploys instead of building features.
The Solution
Introduced strict GitOps with Argo CD, Kustomize, and a CI pipeline that automatically promotes tested images to production.
The Result
Deployment time dropped from 2 hours to <10 minutes, failure rate fell below 5%, and the team gained confidence to ship multiple times per day.
Stack
Kubernetes, Argo CD, Helm, Kustomize, GitHub Actions
Bare‑Metal HA Kubernetes, Self‑Hosted & Fully Documented
The Challenge
Prove that a production‑grade platform doesn't require a cloud provider, and ensure the entire stack is reproducible and transparent.
The Solution
Built a 3‑node HA control plane on legacy hardware using KVM, Ansible, Kubespray. Added MetalLB for load balancing, Cilium for CNI, NFS for storage, and Infisical for secrets.
The Result
Full cluster rebuild automated to <20 minutes (previously 3+ hours). All services run on this platform today, from APIs to a blog CMS.
Stack
Ansible, Kubespray, HA Kubernetes, MetalLB, Cilium, NFS, Infisical
Who I am
Go & Platform Engineer, Certified Kubernetes Administrator and Developer. I build backend services that stay up and infrastructure that stays boring. No magic, no black boxes.
My Philosophy
Software doesn't stop at git push. I deliver systems that are observable, resilient, and fully documented so anyone can own them.
What Sets Me Apart
I don't hide behind managed cloud services. I've built and operate a bare‑metal HA Kubernetes cluster (3‑node control plane, MetalLB, Cilium CNI) from hardware to application. I know exactly what's happening under the hood.
No Black Boxes
Every engagement ends with a runbook that a junior engineer can follow. All code is versioned, all deployments are declarative, and there is zero vendor lock‑in on my personal involvement.
Proven Teacher
- At PerfectStay: trained Java/frontend teams to independently debug and extend Go microservices.
- At ESGI: created the master's curriculum and taught Kubernetes, CI/CD, and Microservices to 200+ students.
Your Benefit
When I leave, your team is fully equipped to run, scale, and extend the system on their own.
Contact
How I Work
Visible progress from week one
Introduction
We left off on a high note. The Kubernetes cluster was alive. But as any engineer knows, a working system is often just the prelude to the next, more interesting problem. While the cluster was technically functional, its architecture had a hidden Achilles' heel: a single point of failure for all incoming traffic.
My mission was clear: eliminate it. The tool for the job was MetalLB, and the task seemed simple. I was wrong.…
You’re in the final interview for a Senior Platform Engineer role. You’re at the whiteboard. The CTO leans forward and asks, "How would you design our observability strategy from scratch?"
You know the tools. You've done this before. Confidently, you dive straight in : "I'd use Prometheus for metrics, Loki for logs because it's cost-effective, OpenTelemetry for tracing...".
The CTO cuts you off. "I don't need a list of tools. I asked for a strategy."
At this moment, you realise that something is missin…
Introduction
For as long as I can remember, I’ve been fascinated by systems—how they work, how they break, and how they can be made better. As a developer, I’ve always sought ways to streamline my workflows and take control of my tools. That’s why I decided to build my own Kubernetes cluster at home. Driven by my passion for systems and coding, I wanted to create a self-owned cloud. I also believe that individual control of digital infrastructures leads to a more balanced world. Setting up eve…
Frequently Asked Questions
Straight answers from someone who builds and operates
My team is all Node/PHP. Isn't introducing Go a huge risk?
I never push a full rewrite. I focus on the 20% of code that causes 80% of the pain - the hot paths that need concurrency and speed. Go runs alongside your existing services, and I train your team so they can own it.
So, are you a developer or a DevOps engineer?
I'm both. I write the Go services and I deploy the cluster they run on. That means I can optimize the whole system, from the HTTP handler to the kernel parameter, without finger‑pointing between teams.
What happens at the end of the engagement?
You get a documented, reproducible system and a team that knows how to operate it. I've trained 200+ engineers and I apply the same method: clear runbooks, recorded walkthroughs, and zero dependence on me.