Skip to content
Sencai.docs

Intelligence & AI

Intelligence is the section of the app where Sencai tries to do some of the analysis and pattern-spotting work for you, instead of leaving you to read dashboards and logs unassisted. It spans several distinct capabilities, built on top of the operational and cost data the rest of the platform already collects - it doesn’t collect any new data of its own, it works from what’s already flowing through Cloud, Fleet, and Operations. This page is a map of what’s here and what’s available to your organization.

What’s in this section

Section titled “What’s in this section”
CapabilityWhat it does
Knowledge graphA map of your organization’s cloud instances, DNS zones, and members, and how they relate
Root cause & correlationsAI-suggested causes for resolved incidents, and cross-service pattern correlation
RecommendationsProactive alerts on cost spikes, incident frequency, and service health
AI governanceYour organization’s AI autonomy level, rate limits, and safety controls

Cost-specific AI features - forecasting, anomaly detection, and right-sizing - are covered in Cloud cost management rather than here, since they’re part of that same workflow.

What’s turned on for your account

Section titled “What’s turned on for your account”

Not every capability in this section runs for every organization:

  • Several of the analysis jobs that populate these screens - knowledge graph, RCA candidates, correlation findings, proactive recommendations - are opt-in, and off by default in most environments. Whether they’re turned on for your organization is something your account team controls, not a setting you’ll find in the app. Each page in this section calls out where that applies.
  • An in-app AI chat assistant (Lumen) exists in the interface, but whether it’s connected to a model for your account depends on a configuration your account team controls too - its presence in the UI doesn’t guarantee it will respond.

If a page in this section looks empty and you expected data, it’s usually because the feature isn’t turned on for your account yet - ask your account team.

Who can use this section

Section titled “Who can use this section”

Reading what’s in Intelligence - the knowledge graph, RCA candidates, correlation findings, recommendations, and your organization’s AI governance settings - is available to any member of your organization, including the Auditor and Viewer roles. Taking an action (accepting or dismissing a suggestion, requesting an autonomy-level upgrade) generally needs Member or higher - each page in this section says exactly what’s required for its own actions, since it isn’t uniform across all of them. Re-seeding the knowledge graph is a separate case: it isn’t an organization-role action at all. It requires Sencai platform-level access, which no organization role - including Owner - grants; see “Platform-level access” in Roles & permissions.

How your data is used

Section titled “How your data is used”

AI features that call out to a language model (recommendations, root-cause suggestions, the assistant) are metered consistently across the platform, so usage always shows up in the same place - see LLM usage for where that shows up. For specifics on data retention, processing, and any subprocessor commitments that apply to your organization, see your Data Processing Agreement rather than this page - that’s the authoritative source, not this documentation.

What’s gated, and by what

Section titled “What’s gated, and by what”

Two independent controls affect what AI capabilities are available to you:

  • Plan tier - the depth of cost-related AI features (anomaly detection, forecasting) scales with your organization’s tier; see Billing & plans.
  • An organization setting, off by default - whether AI agents (external tools using an API token) can take write actions in your account is controlled by a dedicated organization switch, separate from any plan tier. See AI agent access for how that works and who can turn it on.

What’s next

Section titled “What’s next”