Skip to content
ONE3.ai

Platform · Module

Identity & Access

One login across tenants, agent identities, SSO and least-privilege access

Planned · Year 1 (W0)Sold as: IncludedInside the Year-1 boundaryC-2

Identity & Access authenticates people, agents, service accounts, API keys, MCP clients and portal users. It authorises every request through entitlement, licence, roles, scope, field security and conditions.

One login, many tenants

One3 plans a single global user account with a membership per tenant, so a consultant or agency staff member signs in once and switches between clients. Sign-in options include email and password with a breached-password check and lockout, passkeys, Google and Microsoft sign-in and TOTP multi-factor authentication. Access tokens are short-lived and carry the tenant, membership, licence and permission versions, so revoking a session or changing a role takes effect quickly.

Roles that come from the apps

Built-in roles are generated from each app's manifest, with permissions named {app}.{module}.{object}.{action}. Admins clone them into custom roles, create groups and assign roles at tenant, workspace, app, module, department, team or record-set scope. A seat assignment screen prevents licence and role mismatches.

Identities for agents

Every Digital Employee gets an agent identity with an owner membership. Its skill grants are always a subset of the owner's permissions, its workload tokens last 15 minutes or less, and it pauses automatically if the owner leaves. Service accounts and scoped, expiring API keys cover integrations. R4 Actions require step-up authentication.

Enterprise identity

Planned for Wave 2: enterprise SSO over SAML 2.0 or OIDC with domain verification, just-in-time provisioning and break-glass local admins; access reviews with evidence export; consent-based support access with a visible banner; segregation-of-duties rules; and a field-level security screen. Portal identity for customers and partners uses a separate client and cookie domain.

Planned for Wave 3: SCIM 2.0 provisioning, joiner/mover/leaver automation from People, an external AI system register, sign-in risk detection, terms-of-use acknowledgement, rule-based sharing and an access explainer that tells an admin which layer allowed or denied a request.

Features

What Identity & Access includes

Global account, per-tenant membership

One login for many tenants, with a tenant switcher. W0.

Modern sign-in

Email and password with breached-password check and lockout backoff, passkeys, Google and Microsoft sign-in, TOTP MFA. W0.

Short-lived tokens with versioning

5–10 minute access tokens carrying tenant, membership, licence, entitlement and permission versions; session epoch revocation. W0.

Roles from app manifests

Built-in roles generated per app; custom roles, groups and scoped assignments from tenant down to record set. W0–W1.

Agent identities

Agents owned by a membership, skill grants within owner permissions, short workload tokens, pause on owner loss. W0–W1.

Service accounts and API keys

OIDC client credentials and hashed, scoped API keys that expire within a year, with optional IP allow-lists. W0.

Step-up authentication

High-assurance re-authentication within five minutes for R4 Actions. W0.

Session and device management

Users and admins revoke sessions; tenant MFA policy for everyone or admins only. W1.

Record sharing

Record-level viewer and editor sharing, manager chains and portal relations. W1–W2.

Enterprise SSO

SAML 2.0 or OIDC per tenant with domain verification, home-realm discovery, JIT provisioning, enforcement and break-glass admins. W2.

Access reviews and support access

Periodic attest-or-revoke campaigns with evidence export; consent-based, time-boxed support access with banner and audit. W2.

Field-level security

Field permission sets per role and projection masks. W2.

SCIM 2.0 and joiner/mover/leaver

Directory provisioning with group-to-role mapping and deprovisioning within one sync cycle. W3.

Access explainer and sign-in risk

Explains which layer allowed or denied a request; logs sign-ins and flags risky ones. W3.

Workflow

Hire a Digital Employee safely

  1. 1

    Owner creates an agent identity in AI Workforce.

  2. 2

    Skill grants are checked to be within the owner's permissions.

  3. 3

    The agent receives short-lived workload tokens.

  4. 4

    If the owner leaves, the agent is paused automatically.

Workflow

Run an access review

  1. 1

    Admin starts a campaign by role or app.

  2. 2

    The access-review agent flags unused permissions and dormant accounts.

  3. 3

    Reviewers attest or revoke each assignment.

  4. 4

    Evidence is exported for audit.

AI

AI skills and Digital Employees

Access-review agent

Flags unused permissions, dormant accounts, over-privileged agents and SoD risks. Proposals only; revocation is a human decision. W2.

Login anomaly signals

Signals such as API key use from a new country or impossible travel. W2–W3.

Digital Employees that use this module

They act through the same governed Actions as people; risky steps wait for approval.

Frequently asked questions

Do agents get the same access as people?

No. An agent's permissions are always a subset of its human owner's, its tokens are short-lived, and it pauses if the owner leaves.

Is SSO supported?

Enterprise SSO over SAML 2.0 and OIDC is planned for Wave 2 on higher plans, and SCIM provisioning for Wave 3.

Can One3 support staff see my data?

Only with a ticket and your tenant admin's consent, scoped and time-boxed, with a visible banner and an exportable session trail. Planned for Wave 2.

See Identity & Access in your workflow

Tell us about your team and the outcomes you want. We will walk you through the apps, the Digital Employees and the plan for your industry.