AI & Intelligence Overview
TestRelic embeds AI-powered capabilities throughout the platform to help engineering teams diagnose failures faster, understand test quality trends, and get actionable guidance without leaving their workflow.
All AI features require the Growth plan. Starter plan users will see upgrade prompts when accessing AI functionality. See Plans & Billing for details.
AI feature surface
| Feature | Where | What it does |
|---|---|---|
| Ask AI | Dedicated page (/ai) + sidebar | Full-page conversational AI with context attachment, conversation history, and artifact generation |
| Memory | Ask AI composer bar and AI Assistant panel | Injects recent conversation history and long-term user facts into each turn; toggled per user |
| Ask AI Apps | Settings → Integrations → Ask AI Apps | Connect third-party services (Slack, Gmail, Jira, etc.) so the AI can call their actions directly from chat |
| AI Assistant | Floating panel (all pages) | Contextual AI help that understands your current page without switching away |
| AI Insights | Session Workspace → Insights tab | Automated analysis of a test session: defects, user impact, MTTR, Jira linking |
| AI Artifacts | Rendered inside AI chat | Structured outputs: dashboards, reports, test plans, code, data tables, charts, navigation paths |
Context and privacy
When you interact with AI features, the platform can send the following data to the AI model as context:
- Repository names and IDs
- Test run metadata (branch, status, duration — not raw test code)
- Test case names, statuses, and error messages
- Selected context items you attach via the
@mention picker - Prior-session excerpts and long-term facts when Memory is on (scoped to your user account only)
- Tool schemas and action results for any Ask AI Apps you actively attach to a message
No raw source code, environment variables, or secrets are sent to the AI model. Memory context is never shared across user accounts. App data is only included when you explicitly attach the app to a message.
Token usage
AI token usage is tracked per organization and displayed in your subscription metrics. The Growth plan includes 1,000,000 tokens per billing cycle. Usage resets each cycle.
When your token limit is approached, the platform shows a warning. If the limit is exhausted before cycle renewal, AI features are temporarily unavailable until the next billing cycle.
What's next?
- Ask AI — Deep dive into the full-page AI chat experience
- AI Assistant — Learn about the floating assistant
- AI Insights & Artifacts — Session-level insights and structured artifact types