Extensibility & On-Premise (for design partners)

Most teams run Aito as a managed service and never think about where it lives. Some teams have harder requirements: the data cannot leave their walls, the system has to be one they can keep running no matter what happens to the vendor, and the logic they build on top is a real part of their product. This page is for them.

Three things matter here, in order.

1. Run it on your own infrastructure

Aito can run on-premise, inside your own cloud account, your own VPC, or your own datacenter. Your data stays where your compliance boundary already is. Nothing about a prediction, a query, or the data behind it has to cross into infrastructure you do not control.

This is the requirement we hear first from regulated domains like healthcare, finance, and the public sector, where "send your records to a third-party API" is a non-starter before the conversation even reaches accuracy or price. On-premise removes that blocker. You get the engine next to your data, run by you.

2. Keep control of your IP and your continuity

If you are building something an acquirer will one day look at, "we depend on a single outside vendor's database that we cannot take in-house" is a due-diligence problem. The honest answer to it is not a marketing promise. It is control and continuity, written into the agreement:

• A deployment you own and operate. On-premise Aito runs on your side. You are not one API outage or one pricing change away from your product stopping.
• A perpetual license option. The version you deploy stays yours to run, under agreed terms, without an ongoing dependency on us to keep the lights on.
• A source-escrow option. Source held in escrow and released to you on agreed triggers, for example if we were to stop supporting the product, so you can maintain and fork what you depend on. Continuity does not rest on us still being here.

Together these are the real answer to the exit question. Not "trust us," but a deployment you hold and terms that survive us.

3. …and you own what you build on it

Aito is extensible. On top of the engine you can add your own logic: custom scoring, custom functions, and custom column types that run inside Aito, right next to the data, at query time. A worked example is a proprietary scoring model that ranks results by rules only your team knows, computed in-node and cached, so it is both private and fast.

The important part is ownership. What you build on top is yours. The differentiated logic that makes your product yours does not become ours because it runs inside Aito, the same way an application built on a database belongs to the team that wrote it. Aito is the substrate; the edge you build on it stays your edge.

This is the layer we productize last, and deliberately so. The extension API is powerful and still moving. We lock its shape per design-partner agreement rather than freeze a public SDK before we know what real partners need from it. That keeps the surface honest for the people actually building on it.

Where this fits

This page is a preview of where Aito is going for teams with control, continuity, and ownership requirements. It sits outside the standard getting-started path on purpose. The managed v2 API is the front door for most teams, and nothing here changes it. If you are evaluating Aito, start with the Quickstart and the Core Concepts. Come back here when the questions turn to where it runs and what you keep.

Talk to us

There is one next step, and it is a conversation.