Define your entities and relationships once. OrganiseYou turns them into apps, APIs and AI agents that know your data — and only see what they're allowed to.
Platform
One model in the middle. Every door leads to it.
Strings to workflow states, formulas, lookups and auto-numbering — with per-column encryption built in.
Many-to-one and many-to-many with junction fields. Cross-entity lookups resolve wherever you need them.
Layouts and list views configured from the model, not hand-coded per screen.
Workflows with states, activities and transitions. A visual process builder is on its way.
GraphQL, REST with OpenAPI, OData, webhooks and REST importers — generated per tenant.
Permission sets, record sharing and field-level access that every surface — AI included — respects.
How it works
AI models, integrations, your team and your admins all go through the same core. The model decides what exists, how it looks and who may see it.
Access layer · generated per tenant
Tools and actions for AI models, each with a JSON schema and a permission check.
A typed schema per entity, so apps ask for exactly what they need.
Endpoints per tenant with a generated OpenAPI spec.
Queryable feeds per entity for Excel, Power BI and other reporting tools.
Listeners that turn events from other systems into records.
Your entities, typed columns, relations and the records stored in them — validated and permission-checked on every read.
Forms, grids, dashboards, processes and prompt templates — the single source every API, screen and AI prompt is derived from.
Admins shape the model itself: what exists, how it looks and how work flows.
Your team works in screens built from the model — data from Data, layout from Meta — not hand-coded per entity.
AI layer
Your data model already knows what's sensitive. The trust layer uses that knowledge before anything leaves your tenant.
Encrypted and sensitive columns are masked before a prompt is sent — unless a template explicitly allows them.
Every resolved prompt and model response is recorded, so you can see exactly what was asked and answered.
A provider-agnostic gateway for OpenAI, Mistral and Anthropic models, with your own keys and limits.
Agent actions that change data ask for confirmation by default.
Merge fields and grounding go through the same permission checks as a normal read.
Everything is scoped to a tenant. No tenant context means no access — it fails closed.
Roadmap
The foundation runs today. The AI layer and security model are being built in the open — with design partners steering the order. No dates, no vapourware.
Identity, keys, data residency and release control — the things your security and IT teams ask about first.
OIDC and SAML 2.0 — Entra ID, Google Workspace, Okta and more. Verified domains, enforced SSO; SCIM provisioning later.
Protect secrets with a key in your own KMS — Vault/OpenBao, AWS, Azure or GCP. Revoke it and access stops.
Keep your records in your own MongoDB Atlas or Azure DocumentDB cluster.
Flag personal data per column: masked for AI by default, granted per role, with a personal-data access log.
Retention policies, GDPR erasure and anonymisation across records and logs, per-tenant backup and restore.
Sandboxes cloned from production, signed packages from acceptance to live, an environment per release.
Also planned Usage-based plans with metering of records, API calls, AI tokens and storage.
Developers
Tools and actions for AI models, each with a JSON schema and a permission check.
A typed schema per entity, so apps ask for exactly what they need.
Endpoints per tenant with a generated OpenAPI spec.
Queryable feeds per entity for Excel, Power BI and other reporting tools.
Listeners that turn events from other systems into records.
// MCP tool — generated from your model
{
"name": "customer.create",
"description": "Create a Customer record",
"input_schema": {
"type": "object",
"required": ["name", "email"],
"properties": {
"name": { "type": "string" },
"email": { "type": "string", "format": "email" }
}
},
"requires_confirmation": true
}Join the alpha
Request early access and get a personal onboarding session, early access to the AI layer as it ships, and a vote on what we build next.
The core platform runs today, but things will change — sometimes quickly. Expect rough edges, and expect us to listen when you point us in the right direction.
Yes, OrganiseYou is free for design partners during the alpha. Pricing for later stages will be shared well before it applies.
Your metadata — entities, fields, layouts and processes — can be exported as JSON or YAML, and records are available through the normal APIs and exports.
The gateway is provider-agnostic: OpenAI, Mistral and Anthropic models, configured per tenant with your own keys and limits.
Teams that have outgrown spreadsheets but don't want a heavyweight CRM — and developers who want a metadata platform they can extend and integrate.