Every professional I've talked to describes the same morning ritual. Wake up. Grab the phone. Open email. Start scrolling. Try to figure out what's important, what can wait, what's junk. Twenty minutes later, you haven't even gotten out of bed and you're already stressed about your inbox.
The morning briefing exists to replace that ritual with something better.
Instead of opening your inbox and seeing 47 unread messages in random order, you get a single summary — sorted by priority, with draft responses already written for the routine stuff. You read it in 3 minutes. You know exactly what your day looks like. And you haven't opened Gmail once.
Here's exactly how it works, from setup to the briefing hitting your phone at 7am.
You go to the free briefing page and click "Connect Gmail." Google's standard OAuth screen pops up — the same one you see when connecting any app to your Google account. You grant read access. That's it.
We never see your Gmail password. The connection uses Google's official API with a revocable access token. If you ever want to disconnect, you go to your Google account settings and remove access. Takes two clicks.
Once connected, the AI starts processing. It reads your recent email history — the last 30 days — to learn a few things:
This initial processing takes about 24 hours. After that, your first briefing shows up the next morning.
While you sleep, the AI processes every email that came in since your last briefing. For each email, it does three things:
Classifies priority. Every email gets sorted into one of four buckets:
Drafts responses. For every email in the "action required" bucket, the AI drafts a response in your voice. Not a generic template — a specific reply that accounts for the sender, the context, your relationship, and the content of the email. If a client asks about the status of their case, the draft references the case. If opposing counsel proposes dates, the draft checks your calendar.
Generates the briefing. All of this gets compiled into a single, scannable summary document that's delivered to you at the time you choose (default: 7am local time).
Here's what a real briefing looks like. Let's say you're an attorney in Phoenix and it's a Tuesday morning.
42 emails processed overnight. 3 urgent. 14 drafts ready. 9 informational. 16 filed.
URGENT (3)
ACTION REQUIRED (14 drafts ready)
INFORMATIONAL (9)
AUTO-FILED (16)
Read time: 3 minutes. You now know that you need to call Maria Chen about the settlement offer, adjust your calendar for the MSJ hearing change, and get your expert witness section to Tom by Wednesday. The other 39 emails are either handled or irrelevant.
After reading the briefing, you open the approval queue. Those 14 drafts are waiting. Here's what reviewing them looks like:
For the 5 client status updates, you read each draft. They're two to four sentences each, accurately reflecting the current state of each case. You tap "Approve" on all five. Time: 90 seconds.
For the 3 scheduling emails, you glance at the proposed times. They look right — no conflicts. Approve. Time: 45 seconds.
For the 2 discovery follow-ups, you check the proposed delivery dates. One needs to be pushed back a day. You edit that one sentence and approve both. Time: 60 seconds.
For the 2 new inquiries, you read the drafts. They're warm, professional, and include your available consultation times. Approve. Time: 30 seconds.
The bar association and vendor emails: approve, approve. Time: 20 seconds.
Total approval time: under 5 minutes. All 14 emails send immediately upon approval, and your clients get responses before 7:30am — when most attorneys haven't even opened their inbox yet.
Get a personalized AI briefing tomorrow morning. No card required.
Get My Free Briefing →The briefing handles overnight email, but your inbox doesn't stop at 7am. Throughout the day, the AI continues processing in real-time. New emails get classified and drafted as they arrive.
You can handle these two ways:
Either way, you're reviewing and approving pre-written drafts instead of staring at a blank reply box 80+ times a day.
Fair concern. Here's the reality: if you use Gmail, Google's AI already reads your emails for spam filtering, smart reply suggestions, and ad targeting. If you use any email client other than the Gmail web app (Outlook, Apple Mail, Spark, Superhuman), a third party is already accessing your email via API.
AssistantAI uses the same OAuth API access that every other email app uses. We encrypt all data in transit and at rest. We don't store email content long-term — it's processed and discarded. We never use your emails to train AI models. And you can revoke access from your Google settings at any time.
For a deeper look at our security practices, see our security page.
It will, especially in the first week. That's why nothing sends without your approval.
The approval queue is the safety net. You see every draft before it goes out. If the AI misunderstands the context, you edit the draft or write your own response. The AI learns from these corrections — so the same type of mistake is less likely to happen again.
After two weeks of use, most people approve 85-90% of drafts without edits. The remaining 10-15% need minor tweaks — a sentence adjusted here, a date corrected there. It's rare for a draft to be completely off base after the initial learning period.
This is the concern I hear most, and it's the one that's most easily addressed by trying it. The AI doesn't write generic, robotic emails. It writes in your voice — your tone, your word choices, your level of formality. It learns this from your sent email history.
Clients don't notice a difference because there isn't one. The words sound like you because they're modeled on how you actually write. The only thing that changes is the speed of delivery.
And for the emails that truly need your personal touch — the sensitive client conversation, the nuanced negotiation, the difficult news — you write those yourself. The AI flags them as needing your attention and doesn't attempt a draft. You're not losing the personal touch. You're freeing up time so you can actually give personal attention to the emails that deserve it.
You still are. The AI doesn't send anything, delete anything, or archive anything without your approval. It sorts and drafts. You decide. Think of it like having an excellent assistant who opens your mail, sorts it into piles, and writes draft responses — but puts everything on your desk for final sign-off.
The difference between this and doing it yourself is about 2 hours per day.
The first briefing is free. No credit card, no commitment. Connect your Gmail, and you'll have a briefing waiting tomorrow morning.
If you want to keep going — ongoing briefings, real-time drafting, the full approval queue — it's $199/month with a 14-day free trial. Details on the founding member page.
At professional billing rates, $199/month is less than one hour of your time. And it saves you 30-40 hours per month. I'll let you do that math.
I can explain this all day, but the briefing speaks for itself. Connect your Gmail, and tomorrow morning you'll open your phone to a sorted, summarized, drafted inbox instead of 47 unread messages in random order.
If it's not useful, disconnect and forget about it. If it is useful — and I'm pretty confident it will be — you'll wonder why you spent so many mornings scrolling through your inbox the hard way.
Get your free briefing here. Takes 60 seconds.