Daily email admin cut from two hours to 15 minutes with three linked automations
Three connected n8n workflows now sort email, extract tasks and file invoices at the press of a single button.
This builder was losing one to two hours a day to email admin. During the program they designed three connected n8n workflows, writing every custom code node together with AI. Now one button does the sorting, the filing and the follow-ups, and their notes show how much patient debugging a working system takes.
Before
Every day began with one to two hours of email admin. Three to five follow-ups slipped through each week. Each invoice took about 15 minutes to handle by hand, and client information lived in three separate places: Gmail, Drive and a spreadsheet.
“Daily email admin dropped from 1-2 hours to 10-15 minutes.”
In their own words
The turning point
They began working AI-first, using Claude and ChatGPT as architecture partners, code generators and debugging tools. They kept to a strict debugging routine: paste the full workflow JSON into Claude, describe what went wrong, make the smallest fix, test, repeat. Connecting Claude to Notion let it read their databases and prompts directly while they built.
“Debugging followed a strict protocol: paste full workflow JSON into Claude, describe unexpected behavior, implement the smallest fix, test, repeat.”
In their own words
After
They now run three production workflows and six connected Notion databases. One flow takes in email, saves attachments to Drive and links each message to their CRM. A second sorts email into ten categories and pulls out tasks with due dates. A third reads invoice PDFs, records them in the right database and files them by year and type.
What it lets people do
Email admin now takes 10 to 15 minutes a day, and missed follow-ups have dropped to zero. Invoices are handled in one click, and every client's information sits in a single Notion dashboard.
What they learned
Most of the bugs came from n8n losing track of which item belonged to which, so they learned to match data by key and stop trusting item order. They also learned to keep AI prompts in Notion, where they can be changed without touching the workflow, and to check data before it reaches an AI step.
“Build validation gates before AI nodes, not after.”
In their own words
See the work
The Automation Architect: Gmail, Drive, Notion and OpenAI workflow system was built during AI Mastery.
Read the full submission
Written 15 February 2026 Submitted by the storyteller
Every story is written by a Mindvalley student or by someone our teachers have worked with. How we collect and check stories
More from AI Mastery
