Job-Radar
A durable loop direction for measurable job-fit scoring and applications.
AI Engineering · Agentic Workflows · Berlin / Remote
AI agents for people who need systems to work.
I build practical AI tools that reduce friction, increase agency, and help people navigate complex systems — public services, health, democracy, workflows, and mission-driven operations.
agent.loop(read → reason → act → verify)
checkpoint: saved · retry: ready · human: informed
checkpoint: saved · retry: ready · human: informed
Durable loop mindset
01Understand the real human bottleneck.
02Connect tools, data, documents, and LLM reasoning.
03Ship a loop that keeps working when life gets messy.
RoleAI Engineer · AI OperationsBerlin / EU / remote · ready now
ShippedGitLaw · SafeVoice · Recht-MCP · INCA-Loopreal projects, public code, runnable demos
TargetingLangdock · DigitalService · aconium · n8nGovTech · agents · health · workflows
StackPython · FastAPI · LLMs · RAG · MCPClaude Code · Codex · React · TS
The future I want to contribute to
Useful AI should not just make companies faster. It should help people understand, decide, and act — especially when the surrounding system is complex.
Great technology should make people freer, healthier, and more capable. My work focuses on practical AI systems that turn confusion into next steps, documents into decisions, and workflows into reliable loops.
Proof — already shipped
Not just ideas. Runnable systems, public repositories, and demos that show how I think: prototype fast, make the logic visible, and build toward real utility.
AI agents I want to build
Citizen Agent: helps people understand letters, forms, benefits, deadlines, laws, and next steps.
Health Companion: helps patients track symptoms, medication effects, doctor preparation, and emotional load.
Mission Ops Agent: helps NGOs, public-good teams, and health startups automate research, documentation, outreach, and knowledge workflows.
Workflow Builder: connects APIs, tools, documents, and LLMs into useful automations people actually want to use.
Less friction.
More agency.
Better systems
for real lives.
01
Human-first.
I start with the person stuck inside the system.
02
Agentic by default.
I think in loops, tools, checkpoints, and retries.
03
Useful over shiny.
The final question: does it work, and does it help?
Working with me
Curiosity, kindness, courage — and useful weirdness.
I thrive in environments where curiosity, kindness, and courage are alive.
I work best when there is space to explore big ideas, turn them into simple systems, and build something useful for real people. I enjoy working at the intersection of technology, storytelling, ethics, design, and human wellbeing — especially when the goal is to reduce suffering, increase agency, and make complex things feel clear.
I bring the most value when I can move between vision and execution: sensing the deeper pattern, asking better questions, prototyping quickly, learning in public, and improving through feedback.
I care about humane systems, practical AI, consciousness, health, democracy, animals, and tools that help people live with more freedom and dignity.
The environment I thrive in is playful but serious about impact. High-trust. Mission-driven. Honest. Experimental. Warm. A little weird in the best way.
I am at my best when the work feels like service — and when the final question is not only “Does it work?” but also “Does it help?”
Target company map
Berlin and remote teams where AI agents, public-good technology, workflow automation, health, and democratic infrastructure meet.
Langdock
AI adoption, agent workflows, integrations, customer-facing implementation, and real-world automation.
Strong fit — agents, workflows, integrations.
DigitalService Bund
Modern public services, user-centred government software, and digital infrastructure that works for everyone.
Strong mission fit — better public systems.
aconium
Document AI, RAG, MCP integration, and workflow automation for public-sector complexity.
Very strong fit — reads like my project list.
Verwaltungssprung
AI for public administration, digital bureaucracy, trustworthy systems, and better citizen services.
Strong civic-tech fit.
n8n
AI orchestration, API connections, and builder-friendly agentic systems.
Strong stack fit — durable loops and APIs.
voize · Parloa
Humane AI in care and conversational AI at scale. Strong interest; some roles skew senior.
Watchlist / stretch targets.
Current focus: Langdock, Verwaltungssprung, DigitalService Bund, aconium, and n8n. voize and Parloa remain strong watchlist / stretch targets.
What I bring
A practical mix of AI engineering, product thinking, storytelling, empathy, and lived urgency around care, bureaucracy, and systems that should work better.
LLMsRAGAI AgentsDurable LoopsMCPPrompt EngineeringPythonFastAPIClaude CodeCodexWorkflow AutomationProduct ThinkingStorytellingGerman / English / Vietnamese
Let’s build useful AI for complex systems.
I’m looking for Berlin or remote teams building AI agents, workflow automation, healthtech, GovTech, public services, democracy tools, sustainability, or ethical technology.