Junior n8n Automation Implementer (PH-based) — AI/N8N Fanatic • Docker + Postgres

About romea.ai
Singapore-incorporated startup building AI + automation for clinics (US/UK/AU). We run n8n, LLM steps, and self-hosted stacks on Docker + Postgres across 2 redundant prod servers + 1 test. Small PH based senior team, direct feedback, zero drama—accuracy first.

Role (entry-level welcome)
Take rough working flows and turn them into clean, copyable templates. Update all credentials/env for new clients, run tests, fix edge cases, and mark DONE so the core team can focus on new integrations. We like hungry first-job candidates—as long as you already built things in n8n (private projects count).

Location/Hours
PH-based only • Remote


What you’ll do
Duplicate/parameterize n8n workflows; wire up nodes; enforce idempotency + error branches.

Update creds/env per client (GHL often, but also other CRMs/EMRs), run smoke + regression tests.

Add LLM steps (OpenAI/Claude/Groq) with strict JSON outputs and auto-repair loops.

Postgres (must): follow runbooks for setup, roles/permissions, backups/restore (WAL/PITR), migrations, and minor upgrades across prod-A/prod-B/test.

Docker (must): compose files, volumes, healthchecks, rolling restarts, log rotation.

Keep short runbooks and checklists; raise flags early—no surprises in prod.


Must-haves (no exceptions)
You’ve built at least 2 n8n flows (can be hobby/private—show exports or screenshots).

AI/n8n obsessive: you tinker, you test, you care about JSON validity and edge cases.

Solid JS/TS in Function/Function-Item nodes (async/await, fetch/axios).

Postgres + Docker basics via runbooks (you understand what the commands do).

API literacy: OAuth2/API keys, pagination, webhooks, retries/backoff, dedupe.

Communication + accuracy: clear notes, checklists, and reproducible steps. Zero-F-ups mindset.


Nice to have
Python (FastAPI) or Node/Express for tiny services/webhooks.

Redis/queues; Grafana/Prometheus; Linux CLI comfort.

GoHighLevel object model (contacts, opportunities, pipelines, tags).


What success looks like (first 30–45 days)
4–6 client templates cloned, creds updated, tests passed, and handoff notes written.

One basic backup/restore drill on test Postgres following the runbook.

Add idempotency + alerting to at least one flow


Screening (simple)
n8n: links or exports (redact secrets). One paragraph per flow: problem, volume, failure modes.

Function node: 40 lines JS/TS you wrote + what it solved.

Availability (hours/notice), earliest start.

We removed theory questions—show, don’t tell.


Compensation
450hr (18K week, weekly paid)

30–45 day raise path: +50–150/hr when you hit 100% clean handoffs + pass a DB/backup drill.

Optional per-deployment bonus for “template ? creds ? tests ? DONE.”


How to apply (OJPH)
Attach: your n8n exports/screens, Function snippet, and item 6 (rate/availability).
No generic cover letters—show the work.

Skill tags: n8n, JavaScript, TypeScript, Docker, Postgres, API Integration, JSON, Webhooks, Linux, Git, OpenAI/Claude, GoHighLevel, FastAPI.