Complex workflows
Multi-branch logic, loops, merges, and sub-workflows — the orchestration Zapier’s linear model can’t express cleanly.
// n8n track
An n8n developer for production workflows — self-hosted or cloud.
Complex multi-branch workflows, custom and code nodes, self-hosting, error workflows, and the automation Zapier gets too expensive for. 3+ years — n8n is my default for production builds. Same productized pricing: USD, fixed, no surprises.
// capabilities
What I actually do in n8n.
The work that gets billed — not a marketing list. If your gap is here, we can talk concretely.
Self-hosting
Deploy and harden n8n on your own infra (Docker, queue mode, env + secrets) so tasks are effectively free and data stays yours.
Code nodes
Function and Code nodes in JS/Python for anything the standard nodes can’t — transforms, custom API calls, signing.
Custom nodes
Build a custom node when an integration deserves a first-class, reusable block instead of a code step.
Error workflows
Dedicated error workflows, retries, and alerting so failures are caught, logged, and surfaced — not lost.
AI / LLM nodes
OpenAI / Anthropic / agent nodes wired into the flow for classification, extraction, and tool-calling.
Webhooks & APIs
Webhook triggers, polling, and raw HTTP for any REST/GraphQL API — with auth and pagination handled.
Migration & handoff
Move heavy Zapier workflows to n8n to cut task costs, with parity and a clean cutover. Documented, with a Loom.
// packages
The n8n ladder — four ways in.
Spark, Audit, Sprint, or Build — same productized cadence as every track. USD pricing, fixed, no surprises. Bigger or ongoing needs sit one section down.
n8n Spark
$650
One workflow built and shipped — or one failing workflow fixed properly. The smallest way in.
- 1 production workflow — trigger, nodes, and the logic that makes it useful
- Or one failing workflow fixed properly, with an error workflow added
- Run logging + a failure alert so it cannot break quietly
- Short Loom walkthrough + handoff notes
- 7-day post-ship support · paid up front
- Fee credited toward a Sprint if you book within 30 days
n8n Audit
$499
Read-only review of your n8n instance (or your case for one) with a 30-day roadmap.
- 2 × 30-min discovery calls with your ops / eng lead
- Review of existing workflows — structure, failures, gaps
- Self-host vs cloud recommendation for your case
- Map of what’s automated vs what should be
- 30-day improvement roadmap, ranked by ROI and risk
- One quick-win workflow built, ready to import
- 20-min Loom walkthrough on delivery
- $499 credited toward a Sprint or Build within 30 days
n8n Sprint
$2,400
Two to three production workflows shipped end-to-end — the workhorse engagement.
- 2–3 production-ready workflows — pick from:
- – A multi-branch workflow with sub-workflows and merges
- – A code-node integration against an API with no node
- – An error workflow + alerting layer across your automations
- – A webhook-driven pipeline with validation and retries
- One optional AI / agent node step
- Tested failure paths, retries, and run logging
- Documentation + 20-min Loom walkthrough
- 14-day post-ship support · 50/50 billing · Wise / AirTM / wire / PayPal
n8n Build
$4,800
One bigger build — a self-host rollout, a custom node, or a migration off Zapier.
- Choose one shape:
- – A self-hosted n8n rollout (Docker, queue mode, secrets, backups)
- – A custom node for a first-class, reusable integration
- – A Zapier → n8n migration to cut task costs, with parity
- – A multi-system workflow across 3+ apps with shared error handling
- Discovery + scope confirmation before any build
- Security, rate-limit guards, observability, dead-letter pattern
- 30-day post-ship support · one iteration round
- 50/50 billing · Wise / AirTM / wire / PayPal
// bigger + ongoing
Need more than a single build?
Several builds wired into one operation → the Stack bundle. Ongoing help → Automation Care, three tiers from $750/mo. Both are cross-stack and scoped on the main packages page.
// faq
n8n questions, answered.
The ones that actually come up on intro calls.
Self-hosted or n8n Cloud — which should I use?
Depends on volume and data sensitivity. High volume or data you want to keep on your own infra → self-host (tasks effectively free). Low volume / no ops appetite → Cloud. The audit gives you a clear recommendation.
Why n8n over Zapier?
Cost at volume (self-hosted tasks are effectively free), real branching/loops/sub-workflows, code nodes, and data control. For low-volume simple automations Zapier is often fine — I’ll tell you honestly which fits.
Do you build custom nodes?
Yes. When an integration is used a lot, a custom node beats repeated code steps — reusable, typed, and easier for your team to maintain.
Can you wire AI into n8n?
Yes — OpenAI/Anthropic and agent nodes for classification, extraction, drafting, and tool-calling, with the same error handling as the rest of the flow.
How do I know you actually know n8n?
It’s my default for production builds, 3+ years. The $499 audit is the cheapest test, and I’m happy to screen-share real workflows for 20 minutes first.
// next step
Bring me your messiest n8n problem.
The fastest path to a yes-or-no is a 5-minute brief. If it doesn't fit the standard packages, say so — most don't, and most get scoped anyway.