Portfolio
The Portfolio page lists every AI application your organisation has registered with VeriProof. Each card shows the application name, environment (production / staging / sandbox), SDK health status, and the date of the last governed decision.
Filtering and search:
- Use the search bar to filter by application name or tag.
- The Framework chip filter narrows to applications that are mapped to a specific compliance framework (EU AI Act, ISO 42001, NIST AI RMF).
- The Environment filter separates production monitored apps from staging or sandbox experiments.
Application status badges: | Badge | Meaning | |---|---| | 🟢 Healthy | SDK active, decisions flowing, no critical alerts | | 🟡 Degraded | SDK connected but latency elevated or coverage drop > 10% | | 🔴 Critical | Zero decisions in 24 h, or unacknowledged critical alert | | ⚪ Ungoverned | No SDK integration — decisions not being captured |
Bulk actions: Select multiple applications to bulk-assign frameworks, bulk-acknowledge alerts, or export a CSV summary for board reporting.
From each application card, click Open workspace to enter the full Application Workspace for that app.
Portfolio
The Portfolio view is where you oversee all your registered AI applications in one place. Instead of checking applications one by one, you see each application’s governance health side by side — where risk is rising, which applications need attention, and where to act next.
Every role has a portfolio view, but the lens differs. Administrators see the full fleet. Developers see rollout health and SDK coverage. Governance Engineers see policy coverage and maturity. Compliance Officers see evidence readiness and framework gaps. Business Owners see risk and ROI signals. Auditors see a read-only engagement-scoped comparison.
Your Portfolio view opens with the Agent Risk Map and business-language KPI strip. Use the Export action to download a CSV summary suitable for board and leadership reporting.
Your Portfolio view highlights framework coverage gaps and evidence readiness per application. Use the Framework filter to narrow to applications mapped to a specific regulation.
Your Portfolio view is titled Policy Comparison. It shows policy coverage maturity, rule evaluation rates, and which applications need new rules or threshold adjustments. Use it to prioritize where your governance authoring effort will have the most impact.
Your Portfolio view is titled Application Comparison. It shows SDK rollout health, coverage maturity, and policy readiness across all applications. Use it to identify which integrations need work before they can support production-grade governance rules.
Your Portfolio view is read-only and scoped to your active engagement. It shows evidence readiness and framework posture side by side for the applications in scope. Use it to identify which applications need closer review before you submit findings.
You see the full fleet view with all bulk actions: bulk-assign frameworks, bulk-acknowledge alerts, and initiate an emergency stop across multiple applications simultaneously.
KPI Strip
The top of the Portfolio view shows four fleet-level key performance indicators, updated on every page load:
| Metric | Description |
|---|---|
| Fleet governance score | Weighted average governance score across all production applications. Weighting is proportional to session volume so high-traffic applications have greater influence. |
| Applications below target | Count of applications whose rolling 7-day governance score is below their configured target. Click to see the list. |
| High-risk session rate | Percentage of all sessions in the last 30 days classified as HIGH or CRITICAL risk. |
| Human oversight coverage | Percentage of sessions across all applications that carry a human oversight annotation. |
Agent Risk Map
The Agent Risk Map is a visual scatter plot that positions every active application in your fleet on two axes:
- Horizontal axis — session volume (relative traffic load)
- Vertical axis — governance risk score (inverted: higher = more risk)
This makes it immediately obvious which applications are both high-traffic and high-risk — the top-right quadrant. Applications appearing there warrant the most governance investment.
Each application is represented as a bubble. Bubble size indicates the number of open alerts and pending reviews for that application. Clicking a bubble navigates to that application’s workspace.
The risk map updates daily. It is not a real-time view — use Sessions for real-time session investigation.
Action Queue
The Action Queue lists the ten items across your entire fleet with the highest priority scores — a composite of alert severity, time since creation, and the governance impact that resolving the item would have.
Items appear in the queue when:
- An alert rule fires and remains unacknowledged
- A session review has been pending for more than the configured escalation threshold
- A remediation item is past its target date
- A scheduled evidence export failed
Click any item to navigate directly to the relevant workflow. Acknowledging an alert or completing a review removes the item from the queue.
Governance Effectiveness Chart
The Governance Effectiveness Chart shows your fleet-level governance score as a time series over the last 90 days, with the ability to zoom to any 7-, 30-, or 90-day window.
Two reference lines are overlaid:
- Fleet target — derived from the weighted average of all application-level governance score targets
- Industry benchmark — an anonymized percentile band from VeriProof’s operator network, shown where available
Use this chart in monthly governance reviews to demonstrate trend direction — steady improvement, plateau, or regression — rather than focusing on absolute point scores.
Portfolio Q&A
The Portfolio Q&A panel accepts natural-language questions about your AI fleet and returns answers grounded in your session data and governance records. Example queries:
- “Which application had the largest governance score drop in the last 30 days?”
- “How many CRITICAL risk decisions did we have in Q1 vs Q4?”
- “Show me applications with no human oversight annotations in the last 7 days.”
Responses include citations to the specific applications, sessions, or time windows that support the answer.
Portfolio Q&A uses your governance metadata only — it does not have access to session content, prompts, or model responses unless content capture is enabled and you have the appropriate role to view it.