Siirto AI's CTO explains why they built their own entitlement platform instead of using Chargebee or Metron, and why they eventually chose Schematic. The core issue: billing platforms don't handle entitlements, and managing webhooks and databases in-house was too much overhead for a product-led shift.
Chapters
I'm Pietro Fariello. I'm the CTO at Siirto. So essentially what happened is we have and have had for a long time our platform, right? That had its billing and entitlement logic built in the custom web app. When we launched Siirto AI, our MCP server, we needed to get billing and entitlement in-house. Outside of the contractor's grasp. We started looking into billing platforms because that's what we thought we needed. We talked with Metron or Chargebee, realized none of them did what we wanted to because we were actually looking for an entitlement platform. Because the goal was, we don't want to handle databases about entitlements. We don't want to handle webhooks. We don't want to handle the logic of what it means to entitle. Right now it's very much a sales-led business. To convince them to change, you need to give them absolutely everything they have today plus something. We'll be launching, we hope to get it more product-led. The goal is to get it entirely product-led, right? If the product sells itself, I've done my job correctly. I'll tell you what I thought. My assumption, knowing nothing about the industry, was that billing and entitlement would have been the same thing because why would they be different? You're measuring what the users do and what they can do and you bill them for it. My naive mind was like, oh yeah, it's, it's gotta be one platform. In that sense, I think they are gonna converge. Every time we send an invoice, we have to send it to the government, which then sends it to the recipient. All invoices are a contract, which means that every invoice we send has to be like, there is no going back, right? Once it's sent, it's sent. Which has been like the problem for transitioning because we can't create an invoice which we already did send, like when before schematic came about. The other reason we got Schematic was to finally enable self-serve, which we don't have yet. Self-serve is complicated in some ways, but in others it's much, much simpler because you only have a handful of plans. The user manages their billing themselves, whereas trying to migrate our current clients, which were all SaaS-led and so custom plans, everyone has its own thing. I wouldn't describe it that way. I would describe it as essentially the only choice is schematic versus, in my opinion, being dumb and building it. No companies that are putting so much development work into their building platforms, like why? I'm very much of the idea that you shouldn't build anything that isn't your product.
