8:32 AM - Inbox triage
I open the Discovery Call AI dashboard before the coffee finishes brewing. Second month running this, and I'm still checking metrics like I'm waiting for the bottom to drop out. It doesn't. Yet.
The Slack alert from overnight sits in my timeline. Three new signups. Two are from LinkedIn outreach campaigns I had the agent pre-draft last Wednesday. One is a VP Sales at a 120-person marketing automation company in Austin. I flag his signup and make a note to follow up this afternoon.
The dashboard shows:
- MTD revenue booked: $8,420
- Week-to-date closes: 4
- Pipeline value: $34,200 (11 active deals)
- Active users deployed: 23 across 6 customers
My email inbox shows four messages. Two are customer replies from yesterday's sequences. One is a churn notification from Stripe. One is from Priya, my first customer, who apparently loved something enough to send me a note.
I read Priya's email first. "We ran a discovery call yesterday using your agent's outline. The prospect actually said, 'This is the most prepared salesperson I've ever talked to.' We're closing the deal next week. Best $149 I spend every month." I screenshot it and drop it in Slack. These moments matter.
9:45 AM - Review and approve
The agent prepared eight LinkedIn outreach emails for my approval this morning. Cold opens to VP Sales at companies between 80 and 300 people. The AI doesn't send these directly. It queues them for review, and I approve or modify before they go out.
I open the admin UI and scan through them. Most are solid. Specific. They mention something from their LinkedIn profile that signals buying pain. The third one is too pushy. It says "Your team is clearly missing a structured discovery process" without evidence. I edit it to reference an actual post about their company's hiring spree and reframe it as "You're growing fast. Your sales process is probably scaling too. Let's talk about keeping discovery calls high-quality while you scale."
One email is flagged red by the system. The prospect, Carol Reyes at Reyes Family Practice, is already in a different cold sequence we sent two weeks ago. The AI caught the duplicate. I mark that one to pause. I note Carol for a manual follow-up in the next batch. No need to annoy prospects.
The other five I approve. They're out by 10:15.
10:45 AM - A conflict
Slack pings. James at Threshold Tech is reporting something weird. An automated discovery outline the agent generated yesterday had a pricing question that doesn't fit their product. It's B2B enterprise software with variable pricing, and the AI asked "What's your typical price point?" like it was a SaaS with list pricing.
I pull up the campaign in the admin UI. James is right. The agent was trained on patterns from my earlier customers, mostly SaaS. I need to either fix this in the campaign settings or manually override the next outline for Threshold.
I send James a message back: "Got it. I'm fixing the discovery framework for your product type. I'll push an update by end of day and you'll see it in the next agent run. In the meantime, here's a better version of question three." I paste a revised outline.
It's ten minutes of work, but it's my work. The agent is smart. It doesn't know Threshold's pricing model without me telling it.
12:32 PM - Lunch and the metrics check
I grab coffee and sit with the Stripe dashboard open. I need to see who's not using their accounts.
Two customers opened last month but haven't run a single discovery call through the agent. That suggests either they're not using it or they didn't understand how to integrate it into their workflow. I need to know which. I open Gmail and draft an email to Mark at Ascent Sales Solutions: "Hi Mark. I noticed you haven't kicked off a campaign yet. Anything blocking you, or still in setup? Happy to hop on a quick call to walk through it." It's simple. Personal. Not a follow-up sequence.
One more thing. Rachel at Northcrest Capital, who said they'd love to test five seats in early April, just churned. No warning. They're off the monthly plan now. I open Linear and create a ticket: "Why did Northcrest churn. Follow up Monday." I need to know if it was cost, fit, performance, or something else.
Slack shows a demo from the outbound campaign went well this morning. The prospect asked for pricing. That's four closes if this month's rate holds.
2:10 PM - Manual work, escalation mode
David at Granite Partners emails with a billing question. His invoice shows he's been charged twice for the $299 team plan. It's my error. I created his account, set him to the team plan, and then there was a Stripe sync issue that doubled the charge.
I don't have an automated refund flow. This is manual. I open Stripe, find his subscription, issue a manual credit for $299, and email him back: "My mistake on the duplicate charge. I've credited your account for $299 and it should show in your next billing cycle. I'm tightening up the onboarding flow so this doesn't happen again. Thanks for catching it." I tag the incident in Linear: "Billing sync issue on team plan signup. Need to audit other Q2 onboards."
He replies within an hour: "No problem. Really appreciate the quick fix."
These moments matter differently. This is the work that doesn't scale. It's also the work that keeps customers from leaving.
4:15 PM - Pipeline review
I pull up the pipeline spreadsheet I maintain in a shared doc. Eleven deals, currently $34,200 in potential MRR if they all close.
Three deals are stuck. No forward movement in a week. I need to know why. Are prospects waiting on something. Did they ghost. Did they lose budget. I pick up the phone and call two of them. One says they're still deciding but interested. The other, Patricia at Stellar Group, admits they got pulled into a different project and forgot about the discovery demo. We reschedule.
One deal looks solid though. Tom at Velocity Solutions is asking specific technical questions about integration. That's a buying signal. I send him a detailed response and ask about a call next week to talk implementation.
By the end of this, the pipeline is fresher. I don't know which deals will close, but I know which are moving.
5:40 PM - A bug and one more email
I notice a small bug in the admin UI. Campaign status sometimes shows "Running" even though it's paused. It doesn't affect functionality, but it's confusing. I log it in Linear and note: "UI bug: paused campaigns show running status. Low priority, next sprint." It's small, but I document it so it doesn't get lost.
Before I close the dashboard, I send one more manual outreach email. The Austin VP I saw in signup this morning. His name is Marcus Rodriguez, and his LinkedIn shows he just took a new role. I write him: "Marcus, saw your promotion to VP of Sales at Intermark. Congrats. I run a tool that helps teams nail discovery calls, and from what I can see, your team is already strong at this. Probably valuable at scale though. Want to chat 15 minutes next Wednesday?" No hard sell. Just direct.
6:15 PM - The wrap
I close the laptop and reflect. This isn't passive income. This is me, every day, doing the work of an operator. I approve the agent's outreach. I fix the frameworks when they're wrong. I handle the manual billing issue. I chase stalled deals. I chase new ones.
But I couldn't do this without the agent. Eight emails drafted for me this morning, not eight I had to write myself. A framework for discovery calls that works for most customers most of the time. A system that runs on its own and surfaces what I need to review.
Week so far: $8,420 booked. Four closes. One churn I need to investigate. One customer singing my praises. One customer annoyed by a bad outline, now fixed.
I'll take it. Same day next week, probably.