Über mich
I build and stabilise AI agents, automation and full-stack SaaS products. 4+ years, 10+ production systems, from MVP to scalable SaaS for clients in Germany, the UK, India and the US since 2022.
I design the data model and compliance layer first (GDPR, audit logging, data residency), then add the AI layer, so the system holds up under real client load and audit scrutiny.
How I work:
- One issue becomes one pull request, with a review gate before every merge.
- You see running software weekly, not status reports.
- Architecture and compliance are settled before the first line of code.
- If a scope does not hold, I say so before the engagement starts.
Stack: React, Next.js, Node.js, TypeScript, Python, LangChain/LangGraph, PostgreSQL, AWS, Docker, Kubernetes, Terraform.
Send me one paragraph on what you are building and you get my read on scope, the main technical risk and fit within a few hours.
Skills
Expert:in
Fortgeschritten
Portfolio
AI SaaS | Carbon Accounting: Award-Winning Product of the Year
AI Healthcare SaaS: AI WhatsApp & Voice Booking: GDPR, 70% Less Admin
Property Valuation AI Agent: 2-Hour Workups Done in Minutes
Ops Platform Rescue for Accounting Firm: 60% Faster, Audit-Ready
Projekte
GDPR-compliant healthcare scheduling SaaS
Confidential client (healthcare) · Gesundheit und Soziales
Dezember 2024 — heute
Multi-tenant scheduling SaaS for healthcare clinics, built GDPR-first. The data model was designed around consent, role-based access control (RBAC), audit logging and EU data residency before any feature work started. Features include patient booking, staff scheduling, automated reminders and a clinic admin dashboard. Built with Next.js and PostgreSQL, deployed with Docker on AWS. Delivery ran as one issue per pull request with a review gate, and a working demo every week so the client could see real progress and give feedback early.
Next.jsPostgreSQLMulti-agent AI platform for an audit firm
Confidential client (audit) · Steuern und Recht
Dezember 2024 — heute
Multi-agent AI platform that automates routine audit preparation work. LangChain and LangGraph agents handle document intake, evidence matching, working-paper drafting and exception flagging. Every agent output goes to a human reviewer for approval before it is used, and each step is logged so the audit trail stays complete. The PostgreSQL data layer follows GDPR, audit logging and data residency requirements. Deployed with Docker on AWS, with evaluation suites to catch regressions in agent behaviour before release.
PythonCarbon accounting AI SaaS platform
Lnk Technologies · IT, Software und Telekommunikation
Dezember 2021 — heute
Carbon accounting and energy management SaaS with an AI layer, built from PoC to MVP to production. Full-stack build (React, Node.js, TypeScript, Python) with IoT sensor integration, ETL pipelines from NoSQL to a columnar database (query times cut by 80%) and GHG dashboards aligned with ISO standards. Migrating from off-the-shelf tools to a custom platform cut running costs by 80%; IoT reliability improved by 60%. Deployed with Docker on AWS and a self-managed VPS, with monitoring and backups. My role covered architecture, data model, backend, frontend and DevOps. I worked directly with stakeholders on scope and delivery, set up code review and agile delivery for the team, and added an alerting layer for proactive monitoring of the IoT sensors. Today the platform serves live customer reporting on emissions and energy consumption, with the AI layer helping users interpret the data.
React
Berufserfahrungen
Freelance Full Stack and AI Agent Engineer · Vollzeit
Self-employed · Beratung
Dezember 2021 — heute
Independent engineer building and stabilising AI agents, automation and full-stack SaaS products for clients in Germany, the UK, India and the US. I take work from MVP to scalable production, with the data model and compliance layer (GDPR, audit logging, data residency) designed before the AI layer is added. Stack: React, Next.js, Node.js, TypeScript, Python, LangChain and LangGraph, PostgreSQL, AWS, Docker, Kubernetes and Terraform. Delivery model: one issue per pull request with a review gate, working software every week, and scope called out before the engagement starts.
Full Stack Engineer · Vollzeit
Lnk Technologies · IT, Software und Telekommunikation · bis zu 10 Mitarbeiter:innen
Oktober 2022 — Februar 2025
Full stack engineer on a carbon accounting and energy management SaaS, taken from PoC to MVP to production. Built the React, Node.js, TypeScript and Python stack with MQTT-based IoT sensor integration, and ETL pipelines from NoSQL to a columnar database that cut query times by 80%. Replaced off-the-shelf tools with a custom platform, which cut running costs by 80% and improved IoT reliability by 60%. Delivered GHG dashboards aligned with ISO standards. Handled DevOps with Docker on AWS and a self-managed VPS, ran code reviews and worked directly with stakeholders on scope and delivery. Also built CRM features that improved client onboarding efficiency by 40%.
