Loading page…
An AI automation agency stuck on manual DMs bolted on paid ads, an AI voice agent that calls every lead within 5 minutes, and trained closers — $89K in 75 days
227 leads · 69 appointments · 27 deals · ~$3,000 average deal · 40% close rate · $50K + $25K cash collected across two periods
"The bottleneck was the infrastructure, the systems, and the processes." The product was already good — what was broken was that one person was doing DMs, calls and closing for 12 hours a day and none of it compounded.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Before you read anything else: this is a vendor's case study about their own client, presented by the vendor, on a channel with 110 views. Nobody from Infinity Pilot AI speaks. There are no dashboards beyond a screenshot described in passing. The numbers are the agency's numbers about work the agency did, published to sell the same service to you. We have graded it 📎 Creator-Reported and you should read every figure below with that in mind. With that established, the story is coherent and the mechanism is worth understanding. Infinity Pilot AI is an AI automation agency run by a founder named James. They build phone bots, chatbots and email bots for businesses. The presenter is explicit that the product was not the problem: "the product was insane. It's not the product or it's not because of the product that he was not able to sell. The bottleneck was the infrastructure, the systems, and the processes." What James was actually doing was manual outreach — LinkedIn messages, Instagram DMs, cold emails — and it was producing almost nothing. "He was spending a lot of time just doing manual outreaches on LinkedIn, sending some emails on Instagram, and was never fruitful." The more painful detail: he was "stuck on 10 to 12 hours trying to manage everything, trying to be in those DMs, get them on calls, and try to close them." The engagement had a 60-to-90 day window to build something that worked. The reported outcome across 75 days: 227 leads, 69 appointments booked, 27 deals closed, roughly $89,000 in contracted revenue. The first 30 days accounted for about $62,500 contracted with around $50,000 cash collected; the following fifteen days added about $27,000 contracted and $25,000 collected. The irony the presenter points out himself is the useful part: an AI automation agency that had not automated its own front end.
Before the rebuild there was no channel, only effort. James was on LinkedIn messages, Instagram and cold email himself, and it "was never fruitful" — 10 to 12 hours a day spent living in DMs, chasing people onto calls and closing them personally.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself. The exact words are quoted below, with a timestamp that jumps to the second each figure is said.
“And then, in the next 50 day 15 days, we were we contracted around $27,000 in revenue and collected 25K. And we are on the way to actually build it more.”
8:54 · hear it in the video“…was consistent, it brought us a total revenue of around $89,000. And here's what it cost.”
10:24 · hear it in the videoData credibility: 📎 Creator-Reported, and this one needs the strongest caveat on the site. The video is an agency owner (Pursottam Choudhary) presenting a case study about his own client, on his own channel, ending with a pitch to book a call with his team — every number is the vendor's account of his own work, published to sell the same service. Nobody from Infinity Pilot AI speaks or appears. The funnel figures (1,670 opt-ins, 167 leads, 50 appointments, 20 closes, $62,500 contracted / $50,000 collected in month one; 227 leads, 69 appointments, 27 deals, $89,000 over 75 days) are narrated over a Miro board and a briefly-described GoHighLevel snapshot, not shown in auditable detail. The transcript is internally inconsistent at several points — the opt-in count appears as both '1,006 170 viewers' and '160 1,670 opt-ins', and 'with 20 closers we had $62,000 revenue' clearly means 20 closed deals — so auto-captioning has corrupted some figures. Most importantly, the presenter explicitly raises the cost side ('and here's what it cost: ad spend, AI tools, landing pages, VSL, everything') and then states no number at all, so no profitability assessment is possible. The AI voice platform is never named. Deal pricing ($5,000 done-for-you, $2,500 do-it-yourself, ~$3,000 average) is stated but unverified.
The money lands on the closing call rather than after it.
A paid funnel is strictly a machine rather than a flywheel — it produces leads while the ad budget runs and stops when the budget does. What compounds is what the machine leaves behind.
James's flat stretch was the manual outreach period, and the diagnosis is the transferable part of it. Effort was not the missing ingredient, since he was putting in 10 to 12 hours a day across LinkedIn, Instagram and email, and neither was the product: "it's not because of the product that he was not able to sell.
Two things are being described, and it's worth separating them. The thing Infinity Pilot sells is AI automation for businesses — phone bots, chatbots and email bots, packaged into two tiers. A done-for-you system at $5,000 and a do-it-yourself version at $2,500, with the blended average deal value landing around $3,000. The thing that was built for them — and the actual subject of this entry — is a customer acquisition system. It has five stages. Paid ads on Meta drive traffic to an opt-in page. Opt-ins are shown a VSL, a video sales letter, with a booking call to action. Most people don't book from the video, and that gap is where the system lives. Every lead who opts in but doesn't book gets a phone call from an AI voice agent within five minutes. The agent isn't a reminder bot — it holds a conversation, qualifies the lead against criteria, handles objections, and either books a slot on a closer's calendar or actively disqualifies the lead and removes them. Booked appointments enter an automated reminder and nurture sequence to fight no-shows. Human closers take the calls, equipped with an ROI calculator, payment links and an onboarding form so a yes can be converted to cash inside the call. The presenter's summary: "ads bring traffic, our AI and our VA set the appointment, qualifies the prospect, humans close, and we bring in revenue. So this is quality over volume." Our read: the disqualification step is the genuinely clever part, and it's what most people building this get wrong. An AI setter that books everyone floods your closers with garbage and destroys close rate. One that actively removes unqualified leads is the reason the reported 40% close rate is even plausible.
Two packages: $5,000 done-for-you and $2,500 do-it-yourself, averaging about $3,000 per deal. The reported funnel arithmetic, taken directly from the presentation: roughly 1,670 opt-ins produced 167 leads (about 10%), the AI setter called those 167 and booked 50 appointments (a 30% lead-to-booking rate), and closers converted 20 of those 50 (a 40% close rate) for about $62,000 in the first month, of which roughly $50,000 was cash collected. The following fifteen days added about $27,000 contracted and $25,000 collected. Over the full 75 days: 227 leads, 69 appointments, 27 deals, $89,000. We have to flag something here. The transcript's numbers are internally inconsistent at several points — the opt-in figure appears as both "1,006 170 viewers" and "160 1,670 opt-ins", and one sentence reads "with 20 closers, we had $62,000 revenue" where the context clearly means 20 closed deals rather than 20 closers. Auto-captioning is the likely culprit, but it means the funnel figures should be treated as approximate. The ratios (10% opt-in-to-lead, 30% lead-to-booking, 40% close) are stated consistently and are the more reliable part. What is entirely absent is the cost side. The presenter says "and here's what it cost — ad spend, AI tools, landing pages, VSL, everything" and then names no number at all. For a paid-ads-driven funnel with a voice AI layer and commissioned closers, that omission removes any ability to judge whether $89,000 in contracted revenue was profitable. Closer commissions alone typically run 10-20% of collected revenue. Our read: $89,000 contracted over 75 days is a real result for a business that was doing manual DMs. Whether it was a good business is unanswerable from what's shown.
The build started with competitor research rather than creative. "We looked at the industry, where and how the AI automation industry is working, and how these businesses are actually carrying out their processes." The common pattern they found across the category was: Meta ads → opt-in → VSL → human phone setter books the call → closer closes. The common failure they found in that pattern was the setter layer. "The dropping rate was very high. The quality of the calls wasn't that good." Human setters are expensive, inconsistent, slow to respond and hard to manage at small scale — which is precisely the seam an AI voice agent fits into. So the funnel is the industry-standard one with the weakest link replaced. Meta ads carry the traffic, and the presenter emphasises that the creative was built to be "market oriented" rather than generic. The opt-in captures contact details. The VSL does the selling for people ready to book immediately. Then the five-minute rule. Every non-booking opt-in gets an AI call within five minutes — the interval that matters most in inbound lead response, because attention decays fast and whoever calls first usually wins. "Not everyone books immediately, right? So we deployed an AI voice agent that calls the leads within 5 minutes." The reported result of that layer alone: 167 leads called, 50 qualified appointments booked, a 30% lead-to-booking rate with, in the presenter's words, "nobody lifting a finger for doing this work." Automated reminders and nurture then protect the booked appointments, and closers finish the job with the ROI calculator and payment link in hand. Our read: the transferable insight isn't "use AI" — it's the diagnostic method. Map how everyone in your category acquires customers, find the stage with the worst conversion or the highest human cost, and replace only that stage. That's a repeatable way to find automation opportunities in any funnel.
GoHighLevel is the backbone — the presenter refers to a GHL overview in onboarding, GHL dashboard snapshots for the revenue figures, and the CRM and automation setup running inside it. That's consistent with the model: GHL handles the funnel pages, calendar, CRM, pipeline, SMS and email nurture in one place, which is why so many agency operators build on it. Beyond that: Meta for ads, a landing page and VSL, an AI voice agent (the specific provider is never named), automated reminder and nurture sequences, and human closers who are recruited and trained as part of the engagement — "since we hire and vet closers, especially for our done-for-you, we recruit the team for you." The supporting assets built during onboarding: a client agreement, an onboarding recording, sales team SOPs, and tailored sales resources including the ROI calculator closers use on call. Costs: not disclosed. The presenter explicitly gestures at the cost side — "and here's what it cost: ad spend, AI tools, landing pages, VSL, everything" — and then states no figure. This is the single largest gap in the case study, and it is not an accident of editing; a case study that wanted to demonstrate profitability would lead with cost per acquired customer. Our read on what this actually costs to run: Meta ad spend sufficient to generate 1,670 opt-ins, GoHighLevel at roughly $100-500/month depending on tier, an AI voice platform billed per minute of conversation (167 qualification calls plus the calls that don't connect is a meaningful minute count), and closer commission typically at 10-20% of collected revenue. On $89,000 contracted with $75,000 collected, closer commission alone plausibly runs $7,500-15,000. Whether the remainder covers ad spend and the agency's own fee is exactly the question the case study avoids.
Copy the diagnosis before the solution. The presenter is right that the product was never the problem. If you are a competent operator stuck at ten to twelve hours a day of manual outreach, adding more outreach hours is not the fix — the fix is replacing the stage of your funnel that consumes the most human time per unit of output. Copy the competitor funnel audit. Before building anything, map how the best-performing companies in your category acquire customers, then find the stage everyone does badly. That's your intervention point, and it's a repeatable exercise. Copy the five-minute response rule. It is the highest-leverage, lowest-technology idea in the whole case study. Whether an AI agent, a VA or you personally makes the call, calling an inbound lead within five minutes rather than five hours changes contact rate dramatically. Copy the disqualification mandate. Instruct the setter — human or AI — to actively remove unqualified leads from the calendar. A closer's time is the scarcest resource in the system and protecting it is what makes a 40% close rate achievable. Copy the closer toolkit: ROI calculator, payment link, onboarding form, all in hand during the call. A deal that gets a "let me think about it" because there was no way to pay in the moment is a deal you lose to friction, not to price. What to change: start with a human setter, not an AI one. At 167 leads a month, one competent VA calling within five minutes gets you the same 30% booking rate without an unproven voice stack, and gives you the transcripts you need to write the AI agent's script later. Automate the setter once you know exactly what a good qualification call sounds like — automating a conversation you haven't had is how you get an AI that books everyone. Also change the paid-first sequencing. This funnel needs enough ad budget to generate over 1,600 opt-ins before the arithmetic reveals itself. If you don't have that, the same five-minute rule and disqualification discipline applied to organic leads will teach you the funnel far more cheaply.
The evidence is the biggest risk to you as a reader. This is a vendor marketing its own service using its own client's numbers, with no independent verification, no client voice, no ad spend disclosed and no profitability shown. The video ends with a pitch to book a call. Every figure in this entry should be weighted accordingly, and the internally inconsistent numbers in the transcript don't help. Undisclosed unit economics. Without cost per acquisition, this is a revenue story, not a business story. It is entirely possible to generate $89,000 in contracted revenue and lose money doing it, and paid-ads funnels with commissioned closers fail this way routinely. AI voice regulation and reception. Automated outbound calling sits inside a tightening regulatory environment — TCPA in the US, consent requirements elsewhere, and rising state-level rules on AI-generated voice disclosure. A funnel that depends on an AI calling 167 strangers a month has real compliance exposure, and none of it is discussed. Separately, consumer tolerance for AI cold calls is declining, and connect rates that work today may not next year. Ad platform dependency. Meta is the only traffic source. Rising CPMs, an ad account restriction or a creative fatigue cycle takes the whole funnel down at once, and there is no organic or owned channel in the system as described. Closer dependency. Human closers are the conversion layer, they're commissioned, and good ones are scarce and mobile. The system's 40% close rate is a property of the people, not the software. Contracted versus collected. The case study reports both, and the gap is real — $62,500 contracted against $50,000 collected in month one. Payment plans, chargebacks and defaults live in that gap, and on high-ticket agency deals sold to small businesses, they're not rare.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,435 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click