← Back to Home

Service Level Agreements & Support

Last Updated: March 3, 2026 · Version 1.0 (Public Preview)

Severity Definitions

VixPro AI uses a simple two-tier severity model:

SeverityWhat It MeansExamples
CriticalThe platform is down or unable to process alerts — your servers can't be monitoredDashboard unreachable, AI agent not executing tools, WebSocket connections failing, billing preventing access
NormalEverything elseTool output issues, notification formatting, UI bugs, setup questions, feature requests

Platform Availability

Target: 99.5% monthly uptime across all plans.

VixPro AI runs on managed infrastructure from multiple providers. Our availability depends on theirs, and our 99.5% target accounts for that combined dependency.

ComponentProviderSLA Target
DashboardVercel99.5%
AI Agent ContainersFly.io99.5%
Alert ProcessingFly.io99.5%
DatabaseSupabase99.5%

What 99.5% means in practice: Up to approximately 3.6 hours of unplanned downtime per month.

Downtime includes: Dashboard errors lasting more than 5 consecutive minutes, AI containers not accepting alerts for more than 5 minutes, or database unavailability preventing tool execution.

Downtime excludes: Scheduled maintenance (24-hour advance notice), issues with your own servers or monitoring tools, performance degradation that doesn't block core functions, and outages at our infrastructure providers.

Check provider status anytime

  • Fly.io — status.flyio.net
  • Vercel — vercel-status.com
  • Supabase — status.supabase.com
  • Cloudflare — cloudflarestatus.com

AI Agent Performance

The AI agent is the core of VixPro AI. It operates 24/7/365 on all paid plans — no human in the loop required, unless you've designated approval workflows for specific tools.

What We MeasureTarget
Alert acknowledgment (webhook → incident created)Less than 30 seconds
Reactive diagnosis start (alert → first tool execution)Less than 60 seconds
Reactive diagnosis completion (full diagnostic cycle)Less than 10 minutes
Scheduled checks (L1/L2/L3/custom circles)Within 5 minutes of scheduled time
Individual tool execution timeout60 seconds per tool

AI Model Resilience

VixPro AI supports multiple AI providers — Anthropic Claude, Google Gemini, OpenAI, and xAI — through a Bring Your Own Key (BYOK) model. You can designate different providers for different areas of the platform (e.g., one for Reactive alerts, another for Circle groups, and another for Maintenance). If a provider experiences an outage and you have multiple providers configured, you can switch the affected area to an alternative provider to keep your AI agent running.

Proactive Issue Detection

When the AI agent encounters a tool failure or configuration problem during normal operation, it automatically flags the issue and routes it to VixPro AI's engineering queue. Many issues are identified and resolved before you ever need to open a support ticket. You'll be notified when fixes are deployed and your tools are ready to use again.

What Can Affect AI Performance

  • Approval workflows: Tools requiring your approval are paused until granted (5 minutes for NLP and scheduled checks, 10 minutes for reactive alerts, 30 minutes for scheduled maintenance)
  • Alert budget: When the daily alert budget is exceeded, incidents are created in manual mode without AI processing
  • Alert storms: More than 20 alerts in 5 minutes from one server are logged but may have delayed processing
  • Tool permissions: The AI can only use tools you've enabled — broader permissions enable deeper diagnosis

Plan Capacity Limits

 FreePreviewStarterLean
PriceFree$38 one-time$78/month$148/month
Servers0Up to 1Up to 1Up to 3
Team members1126
Stakeholder seats (read-only)126
Alert TriageYesYesYes
Auto RemediationYesYesYes
Priority SupportYesYesYes
API requests/hour11,0001,0002,000
Alert budget/server/day505050
Concurrent reactive agents133

How Support Works

In-App Support Tickets (Primary)

Open a support ticket directly from the VixPro AI dashboard. The support experience works like a conversation — submit your question, include screenshots if helpful, and receive responses right in the dashboard. Every reply triggers a notification bell in the dashboard and sends an email so you never miss an update.

To open a ticket: Click the help icon in the dashboard navigation, select Support → My Tickets, and create a new ticket.

Email

For account issues, billing questions, or if the dashboard is unreachable, email support@vixpro.ai.

AI-Assisted Issue Routing

You don't always need to open a ticket. When the AI agent encounters a tool or platform issue during its normal work, it automatically creates an action item and routes it to the VixPro AI engineering queue. When a fix is deployed, affected tools are re-enabled and you're notified — no ticket required.

Support Tiers

Standard Support — Free Plan

In-app support tickets24/7 submission with screenshot support
Email supportsupport@vixpro.ai
NotificationsDashboard bell + email on every response
AI issue detectionAutomatic — issues flagged and routed proactively
Response timeBest effort, typically within 1–2 business days

Standard Support customers get the full AI platform — same tools, same 24/7 monitoring, same proactive issue detection. The difference from Priority is human response time commitments, not platform capability.

Priority Support — Preview, Starter, and Lean Plans

Everything in Standard, plus guaranteed human response times:

SeverityFirst ResponseUpdatesResolution Target
CriticalWithin 4 business hoursEvery 4 hours until mitigated24 hours
NormalWithin 1 business dayAs progress is madeWithin 5 business days

Business hours: Monday-Friday, 8 AM - 8 PM Pacific Time

First response means a human has read your ticket, triaged the issue, and will either ask for more details or start working on it. It's never an auto-reply.

Critical issues on weekends and holidays: Platform-level critical alerts are monitored outside business hours. Response time may extend to within 8 hours.

Escalation: If you haven't heard back within the stated timeframe, email urgent@vixpro.ai with "ESCALATION" in the subject.

How Issues Get Resolved

VixPro AI uses a structured pipeline to track every issue from detection to resolution. Here's what each status means when you see it on an action item:

StatusWhat It MeansWho Acts Next
PendingIssue identified, awaiting triageVixPro AI
Pending VixPro AI FixThis is a platform-side issue — VixPro AI is working on itVixPro AI
Pending Org ReviewFix deployed — please verify on your endYou
BacklogAccepted as a feature or complex fix, scheduled for a future releaseVixPro AI
ResolvedFixed and confirmedNo action needed

What This Looks Like in Practice

  1. The AI agent runs a scheduled check and a tool fails due to a configuration issue
  2. VixPro AI automatically creates an action item with status Pending VixPro AI Fix
  3. You see it in your dashboard and may receive a support ticket with details
  4. VixPro AI deploys a fix and flips the status to Pending Org Review
  5. You receive a notification that the tool is ready to use again
  6. For complex issues, a backlog item is created and you're kept informed of progress through your support ticket

Many issues follow this path without you needing to do anything — the AI detects the problem, VixPro AI fixes it, and you're notified when it's done.

Maintenance Windows

TypeAdvance NoticeTypical DurationHow Often
Scheduled maintenance24 hours (email + dashboard)Less than 1 hourAs needed, targeting weekends
Emergency maintenanceBest effortVariesRare
AI container updatesNone needed — zero-downtime rolloutNo impactAs released
Tool updatesNone needed — pushed via WebSocketNo impactAs tools are approved

Most updates are deployed with zero downtime. You'll only be notified about maintenance when it could affect your service.

Data & Security

What We ProtectHow
Tool execution historyFull audit trail — parameters, results, and duration logged for every execution
Data in transitTLS 1.2+ encryption on all connections
Your API keyBYOK model — your key is encrypted at rest and never shared
Tool integrityRSA-SHA256 signed manifests verified by both containers and agents
Agent configurationSigned configs with runtime tamper detection — alerts within 5 minutes of tampering
Your dataRow Level Security — your organization’s data is isolated and inaccessible to others

What We Don't Guarantee

We'd rather be upfront than hide behind fine print:

  • AI diagnostic accuracy — The AI uses best-effort diagnosis with the tools and data available. When it can't determine root cause, it escalates to you. This is by design — we'd rather escalate than guess.
  • Your server uptime — VixPro AI monitors and diagnoses your servers but doesn't guarantee their availability.
  • Notification delivery — If Slack, PagerDuty, Twilio, or your monitoring tool has an outage, notifications from VixPro AI may be delayed.
  • API costs — Under BYOK, you control your AI API spend. We provide usage visibility but don't cap your costs.
  • Provider outages — When Fly.io, Vercel, Supabase, or Cloudflare go down, VixPro AI is affected. These events count against our 99.5% target but aren't something we can independently fix.

Future Plans

TierWhat to Expect
Professional (up to 20 servers)Priority Support, uptime credits, dedicated onboarding, expanded limits
Enterprise (20+ servers)Custom SLAs, dedicated infrastructure, data residency options, custom AI model support

Specific commitments will be published when these tiers launch. Existing customers will always be notified before SLA changes take effect.

Public Preview

VixPro AI launched in Public Preview in February 2026. These SLAs are real commitments — not aspirational targets — though they may be refined as we gather operational data. As VixPro AI grows, SLAs will be tightened, never loosened.

Your feedback shapes the product. If something isn't working or you have an idea, open a support ticket — it goes directly to the founder.

Need help? Open a ticket at vixpro.ai/support or email support@vixpro.ai.

Related Policies

These SLAs are part of our legal framework. Please also review: