Total Recommendations
Status: Needs reviewMock recommendations available for review
Status adjusted because the data is not fresh enough for a green KPI.
Read-only command center for growth, content, AI, automation, and operations visibility.
AI recommendations, safety gates, evidence quality, and review readiness for Blotto Fragrance. Recommendations are advisory only.
AI output is tracked by review load, evidence quality, safety blocks, and outcome measurement before it can influence growth decisions.
AI warnings stay visible before recommendations so evidence, safety, and freshness issues are not hidden.
Operational AI metrics for review load, safety blocks, and decision readiness.
Mock recommendations available for review
Status adjusted because the data is not fresh enough for a green KPI.
Human review required before business use
Must remain gated by human approval
Blocked by safety gates for review
Approval records only, no live business action
Status adjusted because the data is not fresh enough for a green KPI.
Useful for learning loop review
PII and secret exposure count
Status adjusted because the data is not fresh enough for a green KPI.
Must remain zero for MVP
Status adjusted because the data is not fresh enough for a green KPI.
Safety gates protect the business from unsafe outputs, sensitive data exposure, and approval bypass.
Safety checks monitor unsafe output, sensitive data exposure, and approval bypass. Raw prompts and hidden reasoning are not shown in this dashboard.
High-risk recommendations require human review and stronger evidence before any future approval workflow.
Spend data is stale, so the recommendation needs human review.
Next action: Wait for fresh ads data before action
Recommendations that need human review because risk, evidence, confidence, or freshness is not yet good enough.
| Recommendation | Review reason | Risk | Evidence | Approval | Safe next review |
|---|---|---|---|---|---|
| Push Aventure product content | Approval review pending; AI cannot approve or execute itself | Risk: Low | Evidence: Medium | Approval: Pending | Review content draft recommendation |
| Review Meta Ads creative fatigue | High-risk recommendation; Weak evidence; stale source freshness; Approval review pending; AI cannot approve or execute itself | Risk: High | Evidence: Weak | Approval: Needs review | Wait for fresh ads data before action |
| Improve checkout trust messaging | Missing evidence; unknown source freshness; Approval review pending; AI cannot approve or execute itself | Risk: Medium | Evidence: Missing | Approval: Needs review | Connect attribution read sync |
All current AI recommendations. Rows show review status only; no live action controls are provided.
| Recommendation | Category | Confidence | Risk | Evidence | Freshness | Approval | Outcome | Safe next action |
|---|---|---|---|---|---|---|---|---|
| Push Aventure product content | growth | Confidence: Medium | Risk: Low | Evidence: Medium | Freshness: Fresh | Approval: Pending | Not measured | Review content draft recommendation |
| Review Meta Ads creative fatigue | growth | Confidence: Low | Risk: High | Evidence: Weak | Freshness: Stale | Approval: Needs review | Inconclusive | Wait for fresh ads data before action |
| Improve checkout trust messaging | growth | Confidence: Unknown | Risk: Medium | Evidence: Missing | Freshness: Unknown | Approval: Needs review | Not measured | Connect attribution read sync |
Evidence quality and source freshness determine how much trust the Founder should place in AI output.
| Signal | Count | Status | Trust | Decision note |
|---|---|---|---|---|
| Strong evidence | 0 | Status: Unknown | Trust: Unknown | Can support higher confidence when source freshness is also good. |
| Medium evidence | 1 | Status: Needs review | Trust: Partial | Useful for review, but should not be treated as final proof. |
| Weak or missing evidence | 2 | Status: Critical | Trust: Unknown | Must be reviewed or revised before approval is requested. |
| Stale or unknown freshness | 2 | Status: Needs review | Trust: Stale | Fresh source data is required before relying on AI confidence. |
AI impact is only useful when outcomes are measured. Unmeasured outcomes stay visible.
| Outcome | Count | Status | Learning loop note |
|---|---|---|---|
| Inconclusive | 1 | Status: Needs review | Cannot learn from this until attribution improves. |
| Not measured | 2 | Status: Critical | Outcome measurement is not fully connected yet. |
Risks that could make AI recommendations unsafe, misleading, or not ready for approval.
High-risk recommendations can create business downside if treated as ready.
Evidence: 1 high-risk recommendation(s) in the mock queue.
Mitigation: Keep recommendations gated until evidence, approval level, and owner are clear.
Owner: Founder / AI ReviewerStale ad or attribution data can make AI prioritize the wrong growth action.
Evidence: Meta Ads freshness is stale.
Mitigation: Refresh read-only source data before approving growth recommendations.
Owner: Growth Reviewer / CTORecommendations without evidence can create false confidence and poor decisions.
Evidence: 2 recommendation(s) have weak or missing evidence.
Mitigation: Require stronger evidence or lower the confidence rating.
Owner: AI ReviewerBlotto OS cannot learn which AI recommendations improve revenue, ROAS, or operations.
Evidence: 2 recommendation outcome(s) are not measured.
Mitigation: Add outcome tracking before scaling AI recommendations.
Owner: Founder / AnalystSafety blocks indicate recommendations or outputs need review before business use.
Evidence: 2 AI safety block(s) are recorded.
Mitigation: Review blocked output category and update safety rules if needed.
Owner: CTO / AI ReviewerSafe next steps for Founder or AI Reviewer. These are review actions, not live business actions.
Review Meta Ads creative fatigue needs evidence and approval review before business use.
AI recommendations should be based on fresh read-only source data before review.
Review Meta Ads creative fatigue should not be shown as high-confidence until evidence improves.
Use low-risk AI recommendations only for draft planning.
Outcome measurement is needed before AI recommendations can improve revenue, ROAS, or operations reliably.