September 11, 2026
SiliconANGLE [reported on September 10, 2026](https://siliconangle.com/2026/09/10/typewise-orchestrates-customer-service-ai-agents/) that Typewise, a no-code AI customer-service platform, has built a system where businesses configure AI agents in plain language -- no code -- to resolve customer requests end to end across email, chat, WhatsApp, and social channels. The headline feature in its Nova offering is the pricing: $2.00 per resolution, pay as you go. A full resolution counts as a full 1.0 charge, a partial hand-off to a human counts as 0.5, and an unresolved request costs nothing. Separately, Typewise launched an AI Supervisor Engine back in February 2026, a multi-agent orchestration layer that coordinates multiple autonomous agents toward customer-service goals, also configured in natural language, deployed with no implementation fees and a free proof-of-value trial. At one customer cited in the report, HealGreen, a Hamburg telemedicine platform, the AI agents reportedly handle 70% of incoming inquiries across email, chat, and WhatsApp, and fully resolve between 75% and 85% of those.
Outcome-based pricing is a genuinely smart correction to how most SaaS support tools have been sold for the last decade. Per-seat licensing charges a business the same amount whether its support tool resolves ten tickets a month or ten thousand. That's a bad deal for a growing business and an even worse deal for a seasonal one -- you pay full freight in the slow months and still hit seat limits in the busy ones. Tying the charge to an actual resolved request, with a discount for partial hand-offs and no charge at all when the agent fails, aligns what a business pays with the value it gets. That's not a marketing trick. It's a real structural improvement, and Typewise deserves credit for building a pricing model that reflects it rather than just talking about outcomes while billing by headcount.
Here's the part worth sitting with: every single resolved support request under this model generates a $2 charge, forever, for as long as the business exists and keeps using it. There's no point at which the meter stops. A business that resolves 5,000 requests a month is paying $10,000 a month, indefinitely, and that number only moves in one direction as the business grows. The HealGreen numbers SiliconANGLE reported -- handling 70% of inquiries, fully resolving 75-85% of those -- describe a tool that's working well. But 'working well' is exactly the condition under which a per-resolution meter becomes expensive fast, because success means more volume, and more volume means a bigger bill. The moment a recurring cost is hardest to escape is precisely the moment the business is thriving and can't afford to rip out the system that's driving that growth.
This is the structural problem outcome-based pricing doesn't touch: it's fairer than per-seat, but it's still rented infrastructure. The business never owns the thing doing the work. It owns a bill that scales with its own success.
Compare that to building the equivalent intake-and-triage workflow directly inside a business's own app -- the kind of FAQ and live-chat intake flow we've walked through building on our own helpdesk and support platform. That approach costs effort once: you configure the routing logic, the escalation rules, the channels it watches. Once it's built, it belongs to the business outright. There's no per-transaction meter running underneath it as request volume climbs from 500 a month to 50,000. The cost of running it doesn't move in lockstep with how well the business is doing. If you want to see the practical differences on paper, our buy vs. build vs. ViibeStack breakdown lays out the same tradeoff in more detail -- and it applies just as directly to AI customer-service tools as it does to CRMs or project trackers.
None of this means Typewise made a mistake. A per-resolution model is the right pricing decision for a company selling access to a hosted AI agent -- it's honest about what it's charging for, which is more than most SaaS vendors can say. The distinction we'd draw is narrower and, we think, more useful: outcome-based pricing fixes the fairness problem with per-seat SaaS. It does not fix the ownership problem. Those are two different problems, and it's worth knowing which one you're solving before you sign up for a bill that grows exactly as fast as your business does.
Sources