The full product
surface.
Ticketing, approvals, SLAs, AI assist, security, and the knowledge base. Each item below is described in the terms you would use to test it during a rollout or a security review.
24
Fine-grained permissions
6 × 12
Automation triggers × actions
3
AI provider adapters: Anthropic, AWS Bedrock, and any OpenAI-compatible (bring your own when self-hosted)
0%
Feature gates between plans
The full ticket model
Custom request types per organization. Per-type custom fields, validation, and approval requirements. Status history with from/to transitions, audited by user.
- Custom intake forms with required-field validation
- Status flow: New → Triaged → In progress → Waiting on requester → Waiting on approval → Resolved → Closed
- Public + internal comments, file attachments, watchers
- Parent / child / blocks / duplicate-of links between requests
- Bulk update + merge
Multi-step with delegation
Approvals are first-class: required by request type, routed to a named approver, delegatable when someone is out of office.
- Approval delegation with date ranges (out-of-office)
- Pause SLA clock while a request is waiting on approval
- Auto-fire automation rules on approval decision
- Audit row on every approval state change
Business hours, pause states, escalations
Per-policy SLA targets that can use business hours, pause when waiting on the requester, and fire multi-tier escalation rules at threshold percentages.
- Per-org weekly schedule + holidays + timezone
- Pause statuses configurable per policy
- Escalation rules: notify · reassign team · bump priority
- Each rule fires at most once per request, so no email storms
Evee: three surfaces, one provider abstraction
Evee is the built-in assistant: suggested replies, ticket summarization, and KB semantic search. Each surface is opt-in per org, audited, and quota-capped; self-hosted brings its own provider.
- 3 provider adapters (Anthropic, AWS Bedrock, and any OpenAI-compatible endpoint)
- Each AI surface switches on per workspace, with a monthly token quota per plan
- Every AI call logged: who used it, on what, and how many tokens
- Turn AI off entirely with one setting (off by default when offline)
The controls procurement asks about
Identity, network scoping, data-subject rights, and audit streaming, each configurable per organization.
- 24 individual permissions, custom roles, team scoping
- Multi-factor authentication + single sign-on (SAML/OIDC) + SCIM provisioning
- Restrict logins to your office or VPN network ranges
- GDPR erasure and data-access export, both with documented APIs
- Stream audit events to Splunk, Datadog, or your SIEM
Versioned articles + hybrid semantic search
Draft → Review → Published workflow with version history. Search matches exact words and meaning together, so "laptop won't turn on" finds the power-troubleshooting article.
- Article versioning with save-by audit
- Categories, tags, search-event tracking
- Articles are indexed for semantic search on publish (bring your own provider when self-hosted)
- Portal AI deflection: suggest KB articles to requesters before they file
- Helpful / not helpful feedback with comment
- Public to org users; embeddable in agent reply composer
The platform plumbing you'd otherwise build yourself
Email, automations, integrations, audit, reporting: the surfaces that turn a ticket model into a working helpdesk.
Email channel
Inbound parsing with thread detection. Outbound via your SMTP.
Automations
6 trigger types × 12 action types with condition matrix.
Multi-team routing
Team-scoped roles. Workload-aware least-busy assignment.
Webhooks + REST API
HMAC-signed webhooks. Rate-limited REST API with key auth.
Macros
Canned responses with variable substitution. Per-org library.
Audit + reporting
CSV export on every list. Audit events stream to your SIEM.
Multi-location / franchise
Run locations as child orgs under an HQ, with a cross-location rollup of open requests, SLA breaches, and agents.
Each connection carries a status stamp: available, an open standard the product implements, or built with the API. See the full inventory on the integrations page →
See it under your own data
Stand up a workspace, import a couple of request types, and route a real inbound email to it. The trial runs on the production product.