Hi all — I build production n8n automations as a software engineer, so the parts that usually stall these projects (custom API calls, custom nodes, LLM/agent logic, self-hosted infra, real error handling) are the parts I’m best at, not where I get stuck.
What I do
- AI-agent & LLM workflows (n8n + Claude/OpenAI + MCP) — support triage, RAG assistants, research agents, content pipelines
- Zapier / Make → n8n migrations, rebuilt cleanly with error handling + retries
- Custom integrations & custom nodes for APIs without a prebuilt node
- Self-hosted n8n (Docker/VPS/Postgres), scaling, security, monitoring
Portfolio — importable workflows you can run yourself: https://automation.levelbrook.com
Workflow JSON + write-ups on GitHub: GitHub - tachyurgy/n8n-automation-portfolio: Production-grade n8n automation workflows — AI agents, lead enrichment, content repurposing. Real, importable JSON by Levelbrook Consulting. · GitHub
Remote and async-friendly. Happy to jump on a short call and scope your use case — reply here or DM.
— Patrick, Levelbrook Consulting