Step by step
Define which email types you get and how they should be sorted/prioritised (categories, urgency, standard replies).
Example promptI want to automate my email inbox with AI. Help me define a system: which categories of email do I typically receive (e.g. enquiries, invoices, newsletters, support)? Suggest 5–7 sensible labels, prioritisation rules and for which types automatic reply drafts make sense. My context: [job/industry].
Write reusable reply templates in your personal tone for the most common email types.
Example promptCreate reusable email reply templates in my tone ([friendly-professional/casual/formal]) for these common cases: [e.g. appointment request, request a quote, decline, follow-up question]. Each template with placeholders for individual details, polite and concise.
Build an automation that reads new emails, categorises them with AI, applies the right label and creates a reply draft.
Example promptWalk me through building an automation in Zapier step by step: the trigger is a new email in [Gmail/Outlook]. An AI step should assign the email to one of my categories [categories], set the matching label and, for the intended types, create a reply draft from my template. Which steps, apps and fields do I need?
Optional: map more complex flows (e.g. save attachments, create tasks, forwarding) visually in Make.
Example promptI want to extend my email workflow in Make: for emails of category [e.g. invoice] the attachment should be saved automatically to [cloud folder] and a task created in [tool]; for [category] a forward to [person]. Describe the scenario with the needed modules and filters.
Frequently asked questions
Is this doable without tech knowledge?
Yes. Tools like Zapier or Make work via building blocks, and the AI describes the setup step by step. Start with a simple flow.
Does the AI send mail automatically?
Usually it only creates drafts that you check before sending. That way you keep control over what goes out.
What about data protection?
Handle confidential content sensitively and check how the providers deal with your data. For sensitive mail, caution is warranted.