An assistant for my inbox,
one that never sleeps in.
For years my inbox owned the first two hours of every day. Now a small, trained bot reads every email, drafts a reply in my voice, lets a second AI check the draft, and either sends it or queues a short list for me. The long threads are handled too. Here is exactly how it works, and the prompts you can use to build your own.
An inbox is a second job nobody applied for.
Most days, the first two hours are eaten by polite replies, soft qualifications, and short answers that anyone could write. This machine reads each email, drafts a reply in your voice, has a second AI check it for sense and tone, and either sends it or hands you the very short list of ones that need a human eye. Here is who it is for, what goes wrong without it, how it works, and what you get back.
Who it’s for
Anyone whose inbox quietly steals the morning. Founders, consultants, coaches, course creators, small agency owners. If you have ever opened your laptop to write one important thing and lost the whole morning to email instead, this is for you.
What goes wrong
Without a system, every email is treated like a small interruption. The day fragments. The important work never starts. By lunch you are tired from typing replies that did not move your business an inch. The inbox owns you, not the other way around.
How the machine works
A bot watches incoming mail, reads each email in context, drafts a reply using the way you actually write, then a second AI checks the draft against rules you set. Good drafts get sent. Borderline ones land in your drafts folder with a tiny note. You scan in seconds instead of writing for hours.
What you get back
Around two hours of your day, back. A clean inbox at the end of every evening. The important conversations still happen, just without you doing the typing. You read, you nudge, the bot writes.
Email used to own my mornings.
There was a season of my business where I dreaded opening my laptop. Not because the work was hard, but because the inbox was full of small, urgent-feeling things. Each one took a minute. Forty of them took the whole morning. By the time I started the actual work I was already tired.
The first version of the fix was a bot that wrote drafts for me. It helped a little. But I still had to open every draft, read it, decide, and click send. The bottleneck moved from writing to reviewing, and the morning still slipped away.
So I added a second AI to the loop. The first bot drafts the reply in my voice. The second one checks the draft against rules I set: is it polite, does it answer the actual question, does it match the way I would write back, is it ready to send. If yes, it goes out. If no, it lands in my drafts folder with a tiny note about what is off.
The change was bigger than I expected. About eight out of ten emails get answered without me. The other two get a draft I can scan in ten seconds. The bot also handles long threads, so when a person replies to its reply, the same loop runs again and the conversation keeps going on its own.
By lunch the inbox is empty, the important conversations are still moving, and I have not written a single email myself. The morning belongs to the work that matters, not to typing the same polite paragraph for the fortieth time.
Proof point: I have documented my actual day on YouTube in A Day In My Life, including the part where the inbox runs itself before I get to my desk.
Two AIs, one inbox, no morning lost
What made this actually work was the second AI. One bot to write, one bot to check. That is the small idea that turned drafts into sent emails and the morning back into mine. Without the checker, every reply still needs a human read. With it, most emails just go.
Train the writer on how you actually reply
Before the first email is answered, the bot is fed a folder of emails you have written. It learns your opening lines, your sign off, your tone with new leads, your tone with old clients, the questions you ask back. From that day on every draft sounds like you, not like a generic assistant. Skip this and every reply comes out slightly off and you cannot trust it to send on its own.
Add a second AI as the checker
The first AI writes. The second AI reads what the first one wrote and checks it against a small set of rules you give it. Did it answer the actual question? Is the tone right for this kind of email? Did it ask anything that should be cleared with you first? If yes to all, the email is cleared to send. If not, it ends up in your drafts folder with a short note. This is the move that lets you trust the machine to send on its own.
Let the bot keep the thread alive
Real conversations are not one reply. They are five. The bot keeps the loop running when the person replies back, reading the new message in the context of everything that came before, drafting again, getting checked again, sending again. By the time you look, a full conversation has happened without you. The deals get done, the meetings get booked, the questions get answered, all while you are doing the work that actually moves the business.
Train the writer, add the checker, let it keep the thread. Do those three in order and your inbox quietly empties itself. Skip any one and you are back to writing replies by hand.
What email used to look like
- Two hours of typing before any real work started
- Forty short, polite, identical replies to similar questions
- Long threads abandoned because the next reply felt like a chore
- Important emails missed because they were buried in the noise
- An empty tank by lunch, and the work day had not started
What email looks like now
- About eight out of ten emails answered without you
- The other two arrive as drafts you scan in ten seconds
- Long threads handled inside the same loop, with full context
- Important emails surface to the top because the noise is gone
- A clean inbox by lunch, and a full tank for the real work
How I train the writer on my own voice
The whole thing depends on the writer sounding like you. This first prompt is what I use to teach the bot how I actually write back, so the drafts feel like mine and I can trust most of them to send on their own.
Train the writer on your voice
I want you to learn how I reply to email. Here is the material: 1. Twenty replies I have written that I am happy with. 2. A short list of opening lines I use. 3. A short list of sign-offs I use. 4. Two phrases I never use that must stay out. Read all of it and tell me, in plain English, the rules of my reply style. Then write me three sample replies to these short scenarios: - A new lead asking what I do - An existing client asking about timing - A cold pitch I do not want Do not invent. Do not stretch. Match the rules you just listed.
Spend twenty minutes here once. Get the voice right and most replies that go out from this day on will sound like you, not like a generic bot.
How I set the rules for the second AI
The checker is the part that turns this from drafts into sent emails. This prompt is what I use to set the second AI up so it knows what to wave through, what to flag, and what to leave alone for me to handle.
Set the checker’s rules
You are the checker. Another AI has written a draft reply to an incoming email. Your only job is to decide if the draft is ready to send as is, or if it should be left as a draft for me to review. Approve the draft for sending only if all of these are true: - It directly answers the question that was asked. - The tone matches the relationship (warm with clients, friendly with leads, polite but firm with cold pitches). - It does not promise a date, a price, or a meeting unless I have already cleared that elsewhere. - It does not raise a new topic that should be cleared with me. If any of those are off, leave the draft in my drafts folder and add one short sentence at the top explaining what is off.
This is the prompt that lets you actually walk away from your inbox. The checker is the difference between a tool that drafts and a tool that runs the day for you.
The 3-minute overview of how this works
Before the build steps, watch this short overview. It’s the exact video from the Automations Made Easy page, and it walks through the mechanics behind machines like this one. 1,000+ students have used these mechanics to save two hours a day, with zero coding.
Want to build an email assistant like this in your business?
1,000+ students. Save 2 hours a day. No coding required. In Automations Made Easy you’ll learn the simple skills to build little machines that read your inbox, draft in your voice, and send the easy ones on their own, just like this one.
What the trained loop quietly does every week
Let me keep this small and believable. I am not going to tell you it makes you a million a year. I will tell you what a modest version really does for a normal week of email.
Say you handle around 40 emails a day, and the assistant clears eight out of ten on its own. That is roughly thirty-two replies a day you never touch. The other eight you scan in ten seconds each. Two hours of your morning, back, every weekday.
Roughly 40 emails a day handled by the assistant
About 8 in 10 replied to without you
Around 2 hours back every weekday, 40 hours a month
Roughly 480 hours a year, without you doing anything once it is set up
Now let it compound. 480 hours a year is twelve full work-weeks you spend on real work instead of typing replies. Over five years that is 60 weeks of your life back, time you can spend on the moves that grow the business or just on yourself. I documented how I run my day on YouTube in A Day In My Life and across my income streams breakdown.
I wrote more about the little habits that keep this loop running cleanly in my Substack, including how I add new rules to the checker without breaking the trained voice.
How I keep long threads alive without losing context
A single reply is easy. A thread of five back and forth is where most bots break. This prompt is how I make sure the assistant reads everything that came before in the conversation, not just the latest message.
Keep the thread alive
When you draft a reply, do not look only at the latest message. Look at the entire thread from the first email to now. Before you draft, list, in two short lines: - What the original ask was - What has already been promised, agreed, or ruled out Then write the reply in my voice. The reply must move the conversation forward, not repeat what has already been said. If the next move is not clear, suggest one specific next step and ask for confirmation. Output the two-line summary first, then the draft reply.
Most assistants reset every reply. This one keeps the whole thread in its head, so the fifth message reads like it remembers the first one.
How I track what the bot does for me
Without a tiny scoreboard you do not really know if the assistant is helping. This last prompt sets up a simple weekly check so I can see how many emails went out without me, how many needed review, and what changed.
The inbox scoreboard
Help me design a one-screen weekly scoreboard for my email assistant. For each week I want to see, at a glance: - How many emails the assistant handled in total - How many it sent on its own (approved by the checker) - How many it left as drafts for me to review - How many were flagged as important and surfaced to me - The rough hours I got back compared to the week before Output the list of numbers to track, then two short paragraphs on how to read them and the single thing to change next week.
With this in place you see the assistant earning its keep. Two hours back a day, week after week, adds up to a different kind of business.
The exact build, step by step
Train the writer on a folder of your own replies
One time setup. You give the bot a folder of twenty emails you have written and are happy with. It reads them and learns your opening lines, your tone with different people, your sign-offs. From that day on every draft sounds like you. Skip this and the assistant will always feel like a stranger writing in your name.
Set the rules for the second AI checker
The checker is the part that turns this from drafts into sent emails. You write a short list of rules that decide what is safe to send and what needs your eye. The bot reads every draft against those rules. Pass equals send. Fail equals draft with a tiny note. Without this step the assistant is just a faster way to write drafts.
Connect the bot to your inbox folder
You point the assistant at a specific folder or a specific label in your email. From then on, anything that lands there is automatically read, drafted, checked, and either sent or left as a draft. Everything else in your inbox is untouched. The bot has clear boundaries and only does what you have told it to do.
Let the bot answer the easy ones on its own
From the moment it is connected, the assistant starts working through the simple replies. New lead asking what you do? Answered. Cold pitch you would not have taken anyway? Politely declined. Question you have answered a hundred times before? Sent. You never see most of these, and that is the whole point.
Scan the short list of drafts in ten seconds each
Every morning a small folder of drafts waits for you. Two or three emails the checker did not feel confident enough to send. Each one comes with a short note from the second AI explaining what is off. You scan, edit a word, click send. The whole review is ten seconds an email, not five minutes.
The bot keeps the threads alive while you work
When someone replies to one of the assistant’s emails, the same loop runs again with the full thread in context. The conversation keeps moving, the meetings get booked, the small questions get answered. By the time you check in, a whole back and forth has happened and the next step is already on your calendar.
An email lands in the watched folder
The assistant starts reading in the background while you carry on
A draft is written in your voice
Trained on twenty of your own replies, with your tone and your sign offs
A second AI checks the draft against your rules
Pass equals send, fail equals draft with a tiny note for your eye
The thread keeps running on its own
Replies come in, the loop runs again, the conversation moves forward
Want to build this in your own business?
Automations Made Easy teaches you, step by step, how to build small assistants like the one you just read about. Zero coding. Plain English. 1,000+ students have already used it to take their mornings back.
The six months after I switched it on
Here is the modest shape of it. The first two weeks are slow because you are still teaching the checker which kinds of emails are safe to send. From week three onwards, the share of emails handled without you climbs steadily and settles around eight in ten.
Share of emails handled without you each week (modest estimate)
Notice the line never spikes to a hundred. That is the point. The last twenty percent is the conversations that should have you in them, and the bot is smart enough to know which ones.
What other students built
I teach the simple skills behind machines like this in Automations Made Easy. Students who built their own version sent back what changed in their first month.
“My mornings are mine again. I used to lose two hours to email and now I open my laptop, scan four drafts, and start the real work.”
“The checker is what changed everything. Before it, every draft needed me. After it, most emails just leave.”
“The bot handles whole threads now. I came back from a long meeting and a discovery call was already booked in my calendar.”
“I trained it on twenty of my own replies and the very next day it sounded like me. People still write back like they are writing to me.”
What’s inside Automations Made Easy
AME isn’t a library of pre-built automations. Every business is slightly different. What’s reusable across all of them is the underlying mechanics: how to set up little machines that listen, write, and follow up while you sleep, and how to wire the pieces together without writing code.
The program walks you through six modules: The Right Tools (the cost-effective, no-code stack I actually use), Task Selection Mastery (which automations are worth building first), Design Secrets (mapping an automation before you build it), Zero to Hero (complete beginner to confident automator), Real-World Application (we build a full automation together, end to end), and Monetization Mastery (turn the skill into a side-business).
It also includes done-for-you templates you import in two clicks, over-the-shoulder training videos, and the same playbook 1,000+ students have used to save two hours a day. No coding required. If you can copy and paste, you can build this.
The email assistant: common questions
Pulled from what readers and Automations Made Easy students ask most.
Will people notice it is not actually me writing?
Not if you train it on a real sample of your own replies. The bot learns your opening lines, your sign-offs, your tone, and the way you answer. The drafts come out sounding like you, and the checker prevents anything off-tone from going out. The few people who do notice tend to like the speed and the clarity more than they care who typed it.
What happens with sensitive or unusual emails?
They land in your drafts folder with a short note from the checker explaining what was off. The bot is conservative by design. If a price has to be quoted, a date has to be promised, or a new topic has been raised, the checker will not send. You read those few drafts every morning in ten seconds each and make the call yourself.
Can it handle long threads or only single replies?
Long threads. When a person replies back, the same loop runs again with the full thread in context. The bot reads everything that came before, drafts the next reply in your voice, and the checker decides whether to send. By the time you look, a real conversation has often already happened on its own.
Does it work with my existing inbox?
Yes. It connects to a specific folder or label in your existing inbox. Everything outside that folder is untouched. You decide which kinds of email are routed to the assistant and which ones stay in your normal inbox for you to handle yourself.
Do I need to code to build one?
No. The whole machine is built with no-code tools by connecting a few friendly steps together. If you can copy and paste, you can build it. That is exactly what Automations Made Easy teaches.
More Money Makers like this one
Built from the same handful of mechanics. Each one captures, converts, or earns money that would have walked out the door.
Build the email assistant yourself, or work with me directly.
If you want to learn the mechanics and build it on your own stack, Automations Made Easy is the playbook. If you’d rather work with me one-on-one to plan it for your specific business, I take a small number of consulting clients each month.
€497 one-time · Lifetime access · 1,000+ students
