Recommendations

Intelligence → Proactive Recommendations (/gravity/analytics/recommendations) is a feed of alerts Sencai’s analysis surfaces on its own, without you having to go looking for a problem first - the platform-wide counterpart to the incident-specific suggestions covered in Root cause analysis & correlations.

What triggers a recommendation

Recommendations are generated at the platform level and shown to the whole organization - there’s no per-user recommendation feed. Each recommendation has a type, reflecting what pattern generated it:

Type	What it’s watching for
Cost spike	Spend rising sharply against your organization’s own baseline
Incident frequency	An unusually high number of incidents in a recent window
Synthetic failure	Repeated failures from your synthetic monitoring checks
Right-sizing	An instance that looks over-provisioned
Security	A security-relevant condition worth surfacing proactively

Reading one

Each entry shows a headline, a priority (low, medium, high, or critical - the list is sorted with the highest priority first), the type badge, a short body explaining what was detected, and when it was detected. Priority scales with how severe the underlying signal is - for example, a larger cost increase or a higher incident count produces a higher priority, not a fixed rule you can configure. There’s no way to filter the list by type or priority from this screen today - you’re looking at the full active list, sorted, every time.

Acting on one

The only action available directly on this screen today is Dismiss, which removes it from your active list. Recommendations don’t currently link through to a one-click fix - where a recommendation points at something actionable (a right-sizing suggestion, for example), follow it to the relevant screen (see Right-sizing) to actually act on it there. There’s no snooze option here the way there is on the cost savings list - dismissing is the only alternative to leaving a recommendation active.

Dismissal & feedback

Dismissing doesn’t ask for a reason, and there’s no separate feedback mechanism tied to a specific recommendation - it simply drops out of your active list. If a recommendation is consistently wrong for your environment (for example, flagging a cost pattern you know is expected seasonal behavior), there’s no in-app way to tune that; raise it with your account team if a whole category of recommendation isn’t useful to your organization.

Note

Proactive recommendations depend on a scheduled analysis job that has to be enabled for your environment - the platform’s design intent is to run it daily. If this list is consistently empty, check with your account team on whether it’s turned on for your account before assuming nothing worth flagging has happened.

NPS feedback

Intelligence → NPS Dashboard (/gravity/analytics/nps) also appears in this section, but it’s worth knowing what it actually is: a dashboard of aggregated Net Promoter Score survey results across Sencai’s whole customer base, restricted to Sencai’s own platform team. If you’re an organization admin, you’ll typically see this screen decline to load rather than show you data - that’s expected, not a bug. It’s listed here for completeness, not because it’s something your organization operates. If you’re asked to rate Sencai and leave a comment somewhere in the product, this dashboard is where that feedback ends up - on Sencai’s side, not yours.

What this doesn’t cover

This screen is separate from cost-specific proactive analysis. Cost anomaly detection and cost forecasting run through their own dedicated screens - see Forecasting & anomalies - rather than surfacing here, even though a cost spike can also produce a recommendation of its own on this page.

What’s next

• Root cause & correlations - a related, incident-focused stream
• Right-sizing - where a right-sizing recommendation actually gets acted on
• Cost allocation & savings - the broader cost-savings recommendation list
• AI governance