Built by a Practitioner, Not a Consultant
Verluna was not founded in a strategy session. It was founded because the same patterns kept repeating across every company, and someone needed to codify what actually works.
The Origin
I am Tolga Oral, based in Berlin. I spent years building marketing automation systems at scale for B2B SaaS companies across DACH. I am not a traditional developer. I am a marketing operations specialist who learned to code because the tools I needed did not exist.
The realization came gradually, then all at once: every company I worked with had the same problems. Disconnected tools that do not talk to each other. Manual processes that someone built in a spreadsheet three years ago and nobody dares to touch. Attribution models that credit the last click and ignore everything that came before it. Lead scoring that lives in someone's head instead of in the system.
The gap was not technology. Every company had HubSpot, or Salesforce, or both. The gap was architecture. Nobody was designing the operating layer between these tools and the business processes they were supposed to serve.
The Discovery
When AI coding tools became production-ready, the scope of what one practitioner could build changed fundamentally. I went from configuring platforms to building systems. From setting up workflows in a GUI to writing semantic matching engines, multi-language translation pipelines, and scoring architectures.
300+ sessions later, what I built was not a collection of automations. It was an operating system: 7 operational domains, 83 specialized skills, 12 external integrations, all running with explicit routing, governance, and memory. The methodology that produced this system is what Verluna offers to every client.
The Thesis
Every enterprise in Europe will run agent-powered operations within three years. The question is whether those agents are duct-taped together by overwhelmed teams or architected into systems that scale and comply.
Verluna exists to build that operating layer. Not to sell AI tools. Not to deliver strategy decks. To design, build, and run the infrastructure between AI capabilities and business processes for European enterprises that need it governed, reliable, and compliant from day one.
What Verluna Is Not
Not a dev shop. We do not build apps to spec. We architect operations.
Not a strategy consultancy. We do not deliver slide decks with recommendations. We build working systems and run them.
Not a staffing agency. One architect with an AI agent workforce delivers the throughput of a 3-to-5-person internal team. You scale operations without scaling payroll.
> Production Numbers
> Philosophy
Code Over Config
Most operations are configured, not built. Configured systems hit ceilings: the platform decides what is possible. Built systems are designed from primitives: you decide what is possible. Verluna builds from primitives.
This does not mean everything needs custom code. It means the architecture is intentional. No-code where it makes sense. Code where configuration fails. Agent-powered where human effort is wasted.
> European Roots
Berlin-based. Built for European data sovereignty and GDPR compliance from day one, not retrofitted for it. Deep understanding of the DACH B2B SaaS landscape: the tools, the buying cycles, the regulatory requirements.
Speaking at the >prompt Developer Conference
April 2026. Topic: AI coding as a non-developer. How a marketing operations specialist went from zero programming background to deploying production AI systems across 300+ sessions. The talk covers the methodology, the tradeoffs, and the specific patterns that made it work.
Not a pitch for AI tools. An honest account of what it takes to build real systems when your background is operations, not engineering.
Six phases to AI-native operations
Watch how operations actually work, not how the org chart says they work.
Break messy reality into bounded domains with clear ownership and routing rules.
Routing, specialization, governance, memory, cadences, observability.
Production systems in weeks using AI as both the development medium and the runtime.
Push every process toward autonomous. Reserve human judgment for decisions that need it.
Turn implicit knowledge into explicit, repeatable frameworks your team owns.
Ready to work together?
Tell me about your operations. What is working, what is not, and what you think agents could change.