Platform · Governance

The answers your security review needs.

Governance is not a page in the trust center. It is whether the platform can enforce a rule you wrote. Infrastructure for AI needs guardrails, and Ward is the warden that enforces them: policy evaluated at query time, for every caller, including the ones that are not people.

Request access Read the docs

What it guarantees

Row-level policy, scoped credentials, and an audit log that covers humans and agents equally.

Row and column policy evaluated at query time, not in the client.

Ward AI drafts access policies from plain English. Nothing binds until a human approves it.

SSO/SAML, SCIM provisioning, RBAC.

Customer-managed keys. Your region, your retention.

One audit log covering every person and every agent in the tenant.

Policy plane

Policy as code.
Not click-admin drift.

The policy plane, kill switch and audit trail your team would otherwise build with a free quarter. Policies are Cedar, written and versioned inside your tenant, and every decision made under them is recorded.

app.getward.ai

A Cedar policy open in Ward’s editor, permitting the viewer role to query the sales.store_sales_daily model when the region is WE, with unit_cost, margin and extended fields marked hidden. — **Policy editor.** Row scope and hidden fields in five lines, with autocomplete over your own models and errors underlined as you type.

Ward’s Policies list for the Acme tenant: eight Cedar policies, seven permitting and one forbidding, each with an inline edit control. — **Policies.** Access is eight Cedar policies you can read in a sitting, not a permissions matrix nobody opens.

Ward’s access log: every authorization decision recorded for the tenant, timestamped to the second in UTC, with outcome, action, resource, actor, and the milliseconds each decision took. — **Access log.** Every authorization decision, allowed or denied, with the actor and the latency. This is the page the security review asks for.

Ward AI drafts. You approve.

Ward AI drafts the Cedar policy from a plain-English brief and nothing binds until you save it. The editor autocompletes actions and entity types against your own models and underlines errors as you type.
Scoped per role, tenant, resource

Finance Agent cannot see labor schedules. Vendor Agent cannot touch shrinkage. US tenant cannot query EU tables. Least-privilege, machine-enforced.
Classifications drive access

Tag a column pii, financial, or operational once. Every agent and query inherits the rule.
Who changed what, on the record

Charter edits, prompt changes, policy updates: logged with name, time, ticket, approver. Diff, roll back, export to your SIEM.
Writes need a named human

Exports, write-backs, schedule pushes. Anything that mutates state gates on a role you define, logged with the approver.
Audit any number on the page

Click a forecast, a margin call, a shrink flag. Ward shows the SQL, source tables, model, parameters, and backtest.

Pause & audit log

One toggle freezes every agent across every tenant, resumes clean, replays nothing. Everything streams to your SIEM as it happens.

Kill switch · JSONL · Splunk · Datadog

Region, tenancy & keys

US stays US, EU stays EU. Single-tenant in your AWS or Azure account on request. KMS or HSM-backed keys; Ward never sees the material.

VPC · data residency · BYOK · CMK · envelope

Insurance & model cards

Cyber and tech E&O on file with an AI rider, COI to procurement in a day. Every number carries a model card with the forecast, the MAPE, and the backtest.

Cyber · tech E&O · AI · forecasts your planners trust

In the console

What this looks like when you actually use it.

Least privilege by default

Read-only credentials, scoped to named schemas. Write access is granted per table and per action. There is no mode in which Ward holds an admin role on your warehouse.

Policy at query time

Row and column policy is evaluated when the query runs, for every caller. A rule enforced in the client is not enforced at all. This applies identically to a person in the console and to an agent.

app.getward.ai