Technical FAQs

For IT teams, developers and technology partners: hosting, security, single sign-on, AI models, MCP, APIs and automation.

For general questions about pricing, set-up and how it works, see all FAQs. For anything not covered here, talk to our team.

Hosting and architecture

Where is our data hosted?

You choose US or EU hosting. With EU hosting, your data and the services that process it stay within the European Union, which matters if you have GDPR or data residency requirements.

How is our data kept separate from other organisations?

Each client runs in its own isolated environment with a dedicated database and storage, so data can't cross from one organisation to another. Every query is scoped to your organisation, and that separation is enforced at infrastructure level, not just in configuration.

Is our content sent to third-party services to be indexed?

No. ReadyIntelligence runs its own search infrastructure, so your documents are indexed and stored on infrastructure we control rather than being sent to a third-party embedding service.

What security certifications does ReadyIntelligence have?

ReadyIntelligence is SOC 2 Type II and ISO 27001 certified, both independently audited. Supporting documentation is available on request for your IT, data protection or procurement team.

Is ReadyIntelligence GDPR-compliant?

ReadyIntelligence is designed to support your GDPR obligations, with permission-aware access controls, source traceability and EU data residency built in, and it is SOC 2 Type II and ISO 27001 certified. Compliance also depends on how your organisation configures and uses the platform, so talk to our team if you need supporting documentation.

Access, identity and sign-in

How does the permissions model work?

ReadyIntelligence operates on a default-deny model — the assistant only sees data sources that have been explicitly configured and permissioned. On top of that, it checks each user's access rights at the point of query, so members and staff only ever receive information they are already authorised to see. You can widen access gradually as your confidence in the platform grows.

Who can see what?

ReadyIntelligence is permission-aware: it applies the access rules you already have, so members-only and staff-only information stays that way.

Which single sign-on options does ReadyIntelligence support?

Two. With OpenID Connect, people sign in through your own identity provider, such as Microsoft Entra ID or Okta, after you register ReadyIntelligence as an application with it. Or your website can pass the assistant a signed token (JWT) for the person already signed in, which ReadyIntelligence verifies before trusting it.

Either way, details your identity provider holds, such as groups or membership type, can decide what each person is allowed to see.

Do our users need another login?

No. Pass the assistant a signed token from your application, or connect your OpenID Connect identity provider, and your users are signed in to it automatically. You choose whether it's for signed-in users only, guests as well, or everyone.

AI models and your data

Which AI models does ReadyIntelligence use?

ReadyIntelligence is model-neutral. It uses leading large language models under the hood and can adopt newer or more capable models as they become available, without you having to change how you've configured your assistants. Answers are always grounded in your own content rather than the model's general knowledge.

Is ReadyIntelligence tied to a specific AI model or provider?

No. The platform is model-neutral by design. It uses leading large language models and can be configured to work with different providers. Model selection is handled at the infrastructure level, not the application level, which means agent behaviour stays consistent even as underlying models are updated or changed.

Can we use our own AI model credentials?

Yes. Clients can supply their own LLM API credentials rather than relying solely on the platform's shared model access. This is useful for organisations with existing enterprise AI agreements, or where transparency over compute costs is a requirement. The platform abstracts model selection, so changing credentials or switching models doesn't require application-level changes.

Does ReadyIntelligence train AI models on my data?

No. Your content and data are used to answer your users' questions, not to train foundation models. Information stays within your permissioned knowledge base, with EU data residency available, and is never fed back into general-purpose model training.

Integration and developer tools

What is MCP, and how does ReadyIntelligence use it?

MCP (Model Context Protocol) is an open standard for connecting AI tools to other systems. ReadyIntelligence uses it in two directions. Other services can be connected to ReadyIntelligence through MCP without custom development. And ReadyIntelligence can itself be used as an MCP server, so your team can work with your knowledge and data from inside Claude and other AI apps, with the same permissions applied as everywhere else.

What systems can ReadyIntelligence connect to?

ReadyIntelligence works with a wide range of content and data: documents, PDFs, presentations, spreadsheets, website pages, cloud storage, knowledge base articles and video, plus live data from connected systems such as your CRM, finance platform and analytics tools. Web crawls can be scheduled to keep content current automatically, and integrations are added without any migration of your existing data.

Can we push content into the knowledge base from our own systems?

Yes. Content can be sent into the knowledge base through an API, so a partner, app or in-house system can keep it up to date automatically. That sits alongside connected sources, scheduled web crawls and MCP connections, so you can use whichever route suits where the content lives.

Can it use data from our own product?

Yes. Connect it to your product's API, or import an OpenAPI description, and it can look things up for the person asking. With single sign-on, each call carries that user's own access, so they only see what they're allowed to.

How much development work is it?

Very little to get started: allow your domain, add one line of script and the assistant is live. Docking it into your layout, adding web components or connecting your API can follow whenever you're ready, and our team can help along the way.

Will it clash with our application's styles?

No. The assistant renders in its own sealed-off layer, so your stylesheets can't reach in and its styles can't leak out. The only thing your CSS touches is the space you give it.

Automation

What can start an automation?

A schedule, such as every Monday morning or the first of the month; a webhook from another system, such as your CRM or a website form; or a person running it when they need it. Large one-off jobs can also be run from an uploaded spreadsheet. Every run is logged, so you can see what started it and what it did.

Can we check what an automation did?

Yes. Every scheduled and triggered run is kept in a full run history, and bulk jobs show live progress with a log of what was created, updated or skipped. You can pause, resume or run an automation whenever you like.

Can we test a bulk job before it runs?

Yes. Upload your spreadsheet, describe what to do with it, then preview and dry-run the job first. Once you're happy with the results, let it run at scale.