AI Transparency

At Vertos AI, we believe in transparency about how our AI systems work, what they can and cannot do, and how your data is used. This document explains our AI capabilities, limitations, and safeguards.

Last Updated: January 2025

1. AI-Powered Features

Our platform uses AI for the following capabilities:

Lead Qualification

Automatically evaluates incoming leads from job boards to determine quality and urgency.

  • Model Type: Rule-based scoring with ML enhancement
  • Purpose: Prioritize high-intent leads for faster response
  • Human Oversight: All leads are visible; scoring is advisory only

Message Personalization

Generates personalized follow-up messages based on customer context and interaction history.

  • Model Type: Template-based with dynamic field insertion
  • Purpose: Save time while maintaining personal touch
  • Human Oversight: Messages can be reviewed/edited before sending

Route Optimization

Calculates optimal routes for technicians considering distance, traffic, and job priorities.

  • Model Type: Optimization algorithms with traffic data integration
  • Purpose: Reduce travel time and fuel costs
  • Human Oversight: Dispatchers can override suggested routes

Smart Job Assignment

Suggests optimal technician assignments based on skills, location, and availability.

  • Model Type: Constraint-based matching algorithm
  • Purpose: Match the right tech to each job
  • Human Oversight: Suggestions only; manual assignment always available

Payment Reminder Timing

Optimizes when to send payment reminders based on customer payment patterns.

  • Model Type: Statistical analysis of payment timing
  • Purpose: Improve collection rates without being intrusive
  • Human Oversight: Reminder schedules can be customized

2. How Our AI Makes Decisions

Input Data

Our AI systems analyze:

  • Lead information (contact details, job descriptions, timing)
  • Customer interaction history (messages sent/received)
  • Job data (location, type, duration, completion status)
  • Technician information (skills, location, schedule)
  • External data (traffic conditions, weather when relevant)

Decision Process

  1. Data is collected from connected systems
  2. Relevant factors are analyzed based on the task
  3. Recommendations or actions are generated
  4. Results are presented for human review when appropriate
  5. Actions requiring confirmation wait for user approval

Automation Levels

Action TypeAutomation Level
Lead capture from job boardsFully automated (data collection only)
Sending initial response messagesAutomated with configurable templates
Route suggestionsSuggestions only; dispatcher confirms
Job assignmentsSuggestions only; manual override available
Invoice generationTriggered by job completion; review available
Payment remindersAutomated sequence; can be paused/stopped

3. Data Usage for AI

Important: Your Data Privacy

Your business data is NOT used to train AI models that serve other customers. Your data is only used to provide services directly to your account.

How Your Data Is Used

  • Service Delivery: Processing your leads, jobs, and invoices
  • Personalization: Learning your preferred response times, message styles
  • Analytics: Generating performance reports for your dashboard
  • Optimization: Improving routes and scheduling within your account

Data Isolation

Each customer's data is logically isolated. AI models do not learn from or have access to other customers' data. Any aggregate insights we develop are derived from anonymized, non-identifiable patterns.

4. Limitations and Known Issues

What Our AI Cannot Do

  • Replace human judgment for complex customer situations
  • Guarantee perfect lead qualification or routing
  • Account for unexpected events (accidents, weather emergencies)
  • Handle highly unusual job types without configuration
  • Make financial decisions on your behalf

Potential Limitations

Our systems may have reduced effectiveness for:

  • New accounts with limited historical data (optimization improves over time)
  • Highly specialized trade services not well-represented in our configurations
  • Regions with limited traffic data coverage
  • Unusual scheduling patterns (e.g., 24/7 emergency-only operations)

When to Use Manual Override

We recommend manual decision-making for:

  • VIP or sensitive customer accounts
  • Complex multi-day projects
  • Situations requiring nuanced communication
  • Emergency or time-critical situations
  • Any time you disagree with an AI suggestion

5. Human Oversight

Your Controls

  • Review Dashboard: See all AI-driven actions and their reasoning
  • Override Capability: Change any AI decision or suggestion
  • Automation Settings: Configure what actions require approval
  • Pause Automations: Temporarily disable any automated sequence
  • Audit Log: Full history of all automated actions

Our Monitoring

  • Continuous performance monitoring of AI systems
  • Regular review of override patterns to improve suggestions
  • Alerts for unusual system behavior
  • Periodic accuracy assessments

6. Compliance Considerations

Our AI systems are designed with regulatory compliance in mind:

  • TCPA/Communication Laws: Automated messages respect opt-out requests and quiet hours
  • Data Protection: AI processing follows our Privacy Policy and applicable data protection laws
  • Human-in-the-Loop: Significant decisions can always be made or reviewed by humans
  • Transparency: This document fulfills our commitment to AI explainability

7. Questions and Feedback

We welcome questions about our AI systems. Contact us at support@vertosai.com

For concerns about specific AI decisions affecting your account, our support team can provide explanations and help adjust settings to better meet your needs.

For privacy-related inquiries about AI data usage, contact privacy@vertosai.com

8. Related Documents