Distributed systems
I make service boundaries, data ownership, and failure paths explicit in systems built with Kafka, asynchronous processing, and multi-tenant services.
Backend and integration
Senior software engineer · distributed systems
From code to production, I stay hands-on.
Hi, I am Fatih. For more than 14 years, I have built software for public services, energy, retail, SaaS, and regulated document exchange. I make architecture decisions close to the code, the data, and the problems that appear in production.
What I work on
The problems I keep meeting in real systems
A technology choice is not a result by itself. Clear boundaries, data ownership, failure behaviour, and the team's ability to operate the system matter just as much as the tools we choose.
Regulated integrations
In Peppol and document flows, I work on delivery together with validation, audit trails, data location, and tenant isolation.
Peppol and compliance
System modernisation
I measure the bottleneck before splitting a monolith. At Mantis, this approach helped one platform grow from about 300 to more than 20,000 concurrent users.
SaaS and cloud
Production ownership
I treat code review, observability, and troubleshooting as part of architecture. The value of a decision becomes clear under real traffic and real failures.
Delivery and operations
Examples from my work
Systems that shaped how I think
I have worked across different industries and technology stacks. The common task has always been to make system boundaries and failure modes visible.
Peppol and document integration platform
I build multi-tenant services that validate, transform, and route documents from ERP and SaaS products to European delivery networks.
SaaS modernisation at Mantis
We separated large enterprise applications into modular SaaS and FaaS components, while scaling one platform beyond 20,000 concurrent users.
Energy and field telemetry
I built store-and-forward services that protected data during connection loss at remote plants, together with high-volume telemetry pipelines.
Latest writing
Notes from production and architecture work
I write about why a decision was made, where it failed, and what it cost in production, using concrete examples whenever possible.
How I Use AI: Operations Are Automated, Decisions Stay With Me
How I turned AI from a chatbot into an operational layer using skills, documentation, independent review, explicit pe...
Read article →Managing Dependencies: Where Should Microservice Boundaries Be Drawn?
Microservices cannot eliminate dependencies. The practical goal is to contain how changes, latency, and failures prop...
Read article →Designing a Dynamic Flow Engine: From Ready-Made Tools to a Custom Execution Model
How a data pipeline becomes a flow, where ready-made tools remain useful, and when a dynamic flow engine becomes a re...
Read article →What Are Kubernetes, Kustomize, Helm Charts, and GitOps?
Kubernetes, Helm, Kustomize, and GitOps manage different layers of delivery. I explain what each one solves and how t...
Read article →How Should We Manage Domain Values in Distributed Systems?
The benefits and limits of enums, strings, registries, and contracts for managing domain codes in microservice and Fa...
Read article →We are at the Beginning of a New Layer of Intelligence
A reflection article on AI, FlyWire, BrainNet, Neuralink and new generation intelligence systems.
Read article →Sovereign Cell Architecture: Regulation-Oriented System Design
Sovereign Cell Architecture: A brief technical note outlining the basic approach and actionable steps on Regulation-D...
Read article →Cloud-Native and Monolit
Real trade-offs between cloud-native approach, platform governance and monolith architectures
Read article →Working on a similar technical problem?
If you want to discuss distributed systems, regulated integrations, or a modernisation decision, you can reach me directly by email.