Workflow Tutorials
How to Build Custom AI Email Agents — No Code, in Minutes
In one continuous demo, this tutorial builds working AI email agents from scratch — no code, just plain-English emails. An email agent in Carly is exactly what it sounds like: an AI assistant with its own inbox address that anyone can email. You describe what it should do in plain English, choose which tools and data it's allowed to touch, and Carly builds it — writing its instructions for you. The demo starts with a fun agent that replies to any email with a custom meme to show how fast an agent comes together, then builds a real support agent that answers questions from a website and CCs you whenever it isn't confident, so nothing gets guessed at. Along the way it handles a taken inbox address live, reviews and fine-tunes the prompt Carly generated, adds the website URL for faster and more grounded answers, and locks the agent down to just web search plus memory — no access to private data — with optional sender authorization so it only replies to approved people. Great for solo founders, small teams, and anyone drowning in repetitive email: support, FAQs, lead response, or internal helpdesks.
Step by step
- 01
See how fast an agent comes together
The demo opens with Meme, a playful agent that replies to any email with a custom meme. It's not the point — it's proof of speed. In seconds you can see the whole shape of an email agent: it has its own inbox address, anyone can email it, and it acts on whatever comes in.
- 02
Describe your agent in plain English
To build your own, you don't write code or wire up steps. You describe what the agent should do in a sentence or two, and Carly turns that into a working agent — including writing its instructions for you. This one is a customer-support agent that answers questions about your business.
- 03
Answer questions from your website — and CC yourself when unsure
Point the agent at your website so its answers are grounded in your real content instead of guessed. Just as important: tell it to CC you whenever it isn't confident. That way a human sees the edge cases and the agent never bluffs an answer to a customer.
- 04
Name it and handle a taken inbox address
Every agent gets its own inbox address. When the obvious choice (like hello@) is already taken, you just pick another — the demo hits this live and resolves it on the spot, so you know what to expect when you set yours up.
- 05
Test it and review the prompt Carly wrote
Send the agent a test email and watch it reply. Then open the instructions Carly generated and fine-tune them — adding your website URL and tightening the wording makes answers faster, more grounded, and more on-brand.
- 06
Lock it down before you deploy
Restrict the agent to only the tools it needs — here, web search plus memory, with no access to your private data. For extra control, turn on sender authorization so it only replies to approved people. This is the step that makes an agent safe enough to actually put in front of customers.
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
See what people say
"Before Carly, I relied on a Calendly link, but the whole process felt impersonal and not very professional. Carly changed that by handling all the back-and-forth, so I'm no longer stuck in endless email threads trying to line up schedules.
Now Carly reaches out to candidates, shares my real-time availability, lets them pick a slot, then sends a Zoom link and drops it straight into my calendar. She sends reminders to both of us before each call, which has significantly reduced no-shows and last-minute confusion.
On top of scheduling, Carly acts like a full executive assistant, sending me my schedule the night before so I can prepare for each call. It reminds me of the old x.ai assistant, but Carly is noticeably smarter, faster, and better suited to my healthcare recruitment business."