Product
React and TypeScript operator UX, product flows, usage controls and billing surfaces.
01 / Featured system
GetAutomata · 2025 to present
I built the product and the system it runs on.
GetAutomata is an AI automation product engineered across the interface, application layer, execution path, data boundary and production infrastructure.
Visit GetAutomataApplication
Node.js and Express REST APIs, Auth0/JWT, and a product boundary that keeps infrastructure internal.
Execution
A custom isolated runner, Redis/BullMQ queues and custom workers for controlled asynchronous execution.
Data
PostgreSQL and Supabase with tenant scoped persistence, Row Level Security and durable execution records.
AI & agents
LLM orchestration, n8n automation and agents with explicit roles, bounded tools and approval gates.
Infrastructure
Docker, Kubernetes/k3s, Cloudflare, Hetzner, Railway, CI/CD and production observability.
02 / Execution architecture
Systems, not demos
I own the full execution path.
I prefer owning the complete path from user action to production execution, including the controls, queues and failure paths between them.
- 01User
- 02Product UI
- 03Product API
- 04Auth / economic control
- 05Runner
- 06Queue
- 07Workers
- 08AI / automation execution
- 09Persistence
- 10Result
I built the runner.Bounded, isolated and protected against replay.
I built the workers.Queues, retries, recovery and operational control.
I built the infrastructure.Deployment, service boundaries and observability.
03 / AI & agent engineering
Agents with explicit responsibilities. Not chatbot demos.
I design AI systems as software systems: explicit responsibilities, controlled tool access, deterministic infrastructure around probabilistic models, and human approval where actions carry operational risk.
4built internal agents
3engineering agents
1marketing outreach agent
Explicit roles
Each agent has a defined purpose, scope and operating contract.
Controlled capabilities
Tools and credentials sit behind scoped boundaries instead of inside prompts.
Deterministic shell
Queues, budgets, idempotency and persistence surround model execution.
Human authority
Operationally risky actions require explicit approval and auditable evidence.
04 / Production engineering
Shipping the feature is only part of the job.
I design the deployment, security boundaries, observability and failure paths that keep it operating.
Execution reliability
Separated orchestration and tenant queues, bounded concurrency, retry policy, dead letter recovery and graceful shutdown.
Security boundaries
Auth0/JWT, scoped execution authority, containers without root access, networks that deny access by default, RLS and verified webhooks.
Observability
Prometheus compatible metrics, queue and runtime health, Grafana dashboards, alerts and structured logs with trace context.
Delivery control
Docker and k3s delivery, SAST and dependency checks, immutable artifacts, releases tied to exact commits and staged promotion.
Production stack Docker · Kubernetes/k3s · Cloudflare · Hetzner · Railway · PostgreSQL/Supabase · Redis/BullMQ · Auth0/JWT · Stripe · Prometheus · Grafana
05 / Verified engineering evidence
Verified, not embellished.
Current repository evidence replaces historical CV metrics. Counts below come from the September 2026 source snapshot.
CI/CD stages
current backup definition
pipeline jobs
current backup definition
test definitions
statically enumerated
test files
code files with test names
built internal agents
3 engineering · 1 outreach
Architecture proof includes a custom isolated runner, two worker roles on separate queues, controlled execution scopes, dead letter recovery, release gates tied to exact commits and private k3s workloads.
06 / Industrial systems background
From commissioning systems worldwide to building AI products.
Before building AI products, I commissioned industrial automation systems across four continents. That taught me to diagnose the whole system, work under operational pressure, communicate directly with customers, test against measurable acceptance criteria and remain responsible until the system works.
At SSI Schäfer and Louis Schierholz, I worked from installation and PLC configuration through performance testing, technician training, customer acceptance and handover.
4continents of technical delivery
98.5%required acceptance threshold
80%+assignments worked independently
June 2025 to present
Founder / AI Product & Full Stack Engineer
Trendiva Lux / GetAutomata · Düren, Germany
Architecture, product engineering, execution systems, infrastructure and daily operations.
February to December 2024
Automation / Commissioning Engineer · Power&Free 350
Louis Schierholz GmbH · Bremen, Germany
On site coordination, PLC configuration, acceptance testing, customer training and handover.
June 2020 to May 2022
Web Project Manager & Frontend Developer
Remote
Frontend delivery, project coordination and customer facing execution in a remote environment.
May 2017 to April 2020
Onsite Manager / Commissioning Engineer · CUBY
SSI Schäfer d.o.o. · Global deployments
International commissioning, performance testing, fault diagnosis, technician leadership and customer acceptance.
April 2016 to April 2017
IT Systems Technician
Zaquensis / Computacenter · Kerpen, Germany
Hardware and system configuration, updates, maintenance and technical fault diagnosis.
Education Matura, Gymnasium Ptuj · Architecture studies, University of Maribor, 2012 to 2014 (not completed).
Languages English at near native level · German professional / negotiation level · Slovenian native.