In 64 days on Supplo, PVAPins handled 6,512 customer conversations across live chat — and let AI close 4 out of every 5 of them, without a human ever touching the thread. Here is exactly how the numbers played out.
PVAPins sells phone-number rentals for SMS verification — a service where customers come and go quickly, ask the same handful of questions, and expect answers in seconds. "Is this number available in Brazil?" "My code didn't arrive." "How do I top up?" "Do you accept USDT?" Multiply that by hundreds of new visitors a day, in a dozen languages, and you have a support load that no three-person team can sustain by hand.
Before Supplo, every conversation routed to a live human, even the ones that were answered verbatim in the FAQ. That meant slow first-response times during peak hours, late-night queues, and a support team that spent most of its day copy-pasting the same five replies. Crucially, the kinds of questions PVAPins gets — order status, country availability, payment-method help — are almost perfectly suited to AI automation if the system can read the knowledge base and learn from past resolutions. They just needed the system.
PVAPins picked Supplo for three reasons that map directly to their economics:
Setup took an afternoon: paste the widget snippet, import six core knowledge-base articles (the existing help-centre content), turn on AI auto-reply. That was it.
Here is what the Supplo dashboard shows for PVAPins as of today — pulled directly from production:
Intercom's public pricing charges $0.99 per AI resolution on top of per-seat fees. Supplo charges $0.04 per AI resolution above plan limits — and at PVAPins' volume, all 4,931 resolutions are covered inside their flat $199/mo plan. Here is the side-by-side for the 2-month period:
| Cost item (2 months) | Intercom equivalent | Supplo (actual) |
|---|---|---|
| Platform fees | $74/seat × 3 seats × 2 mo = $444 | $199/mo × 2 = $398 |
| AI resolutions (4,931) | 4,931 × $0.99 = $4,882 | Included in plan ($197.24 at marginal rate) |
| Total spend | ~$5,326 | $398 |
| Net savings | — | $4,928 in 2 months |
Even calculating only the AI-resolution line, PVAPins saved $4,685 in 64 days. Annualised at the current run rate, that is roughly $28,000 a year kept inside the business instead of paid out per-resolution.
The trick is that Supplo's AI does not just read the knowledge base — it builds its own. PVAPins started with 6 hand-written knowledge-base articles. Today, Supplo has generated 4,764 self-learned knowledge chunks by extracting reusable answers from every conversation a human closed. The AI now has nearly 800× more material to draw on than what the team manually wrote.
That is why the resolution rate keeps climbing month-over-month: every human-handled ticket teaches the system how to handle the next one autonomously. The 19% of conversations that still need a human are the genuinely novel ones — refund edge cases, account-specific issues, the things that should have a human on the other end.
At the current run rate, PVAPins is on track to handle roughly 36,000 conversations per year with ~28,500 AI-resolved automatically. The team's three humans now focus on payment disputes, custom country requests, and partner onboarding — the high-value work that grows the business. The repetitive 81% runs itself.
The annualised savings vs Intercom: roughly $28,000/year in resolution fees alone, without counting the seat-based savings. For a small team, that is one and a half months of payroll — or, more practically, the budget for a fourth hire who actually moves the business forward.
14-day free trial. No credit card. Set up in 5 minutes.