Design an n8n workflow before adding AI
How to map inputs, outputs, systems, validations, retries, logs, and review before using a model.
Read runbookRUNBOOKS
Simple playbooks for production n8n workflows.
These guides explain how I think about automation: stable workflow first, AI only where useful, review where risk matters, and handover so the team can operate it.
AI inbox triage with human review
A practical pattern for sorting emails, drafting replies, routing exceptions, and keeping control.
Read runbookTender and specification review workflow
How to turn tender PDFs and specs into extracted requirements, risk notes, and review tasks.
Read runbookCRM follow-up automation in n8n
How to keep follow-ups, meeting notes, reminders, and owners from depending on memory.
Read runbookEvaluate AI outputs inside n8n
Simple checks for confidence, missing fields, source support, policy rules, and human approval.
Read runbookWant this built for your process?
Use a runbook to understand the pattern, then request an audit to map it to your tools and data.
Request an audit