AutoLead

AI Agents

A deliberately lean agent architecture

Four agents, each with one mission. They share the same signals and the same loop: signals → reasoning → decision → action → outcome → learning.

SignalsReasoningDecisionActionOutcomeLearning

Fleet Optimisation Agent

Maximise fleet utilisation and asset productivity.

Active

Reasoning runs today

48

Recommendations

26

Acceptance

73%

Realised value

AED 486K

Reasons across

DemandAvailabilityLocationVehicle classIdle timeMaintenanceRental durationFleet movement

Produces

  • Demand Forecast
  • Utilisation Forecast
  • Fleet Reallocation Recommendations
  • Idle Fleet Opportunities
  • Asset Productivity
  • Capacity Risks

Customer Journey Agent

Deliver proactive, personalised customer experiences throughout the rental lifecycle.

Active

Reasoning runs today

2,140

Recommendations

884

Acceptance

69%

Realised value

AED 372K

Reasons across

Journey stageSentimentRental historyPreferencesAvailabilityPolicy knowledgeConversation context

Produces

  • Next Best Action
  • Extension Probability
  • Upgrade Propensity
  • Churn Risk
  • Customer Sentiment
  • CLV
  • Human Escalations

Escalates to a human with full conversation context on complaints, damage disputes and legal or fine matters.

Revenue & Pricing Agent

Maximise revenue and margin from available fleet and demand.

Active

Reasoning runs today

36

Recommendations

21

Acceptance

57%

Realised value

AED 298K

Reasons across

Demand forecastRate historyCompetitor ratesChannel mixAvailabilityBooking paceMargin

Produces

  • Pricing Recommendations
  • Demand Opportunities
  • Upgrade Recommendations
  • Add-on Recommendations
  • Channel Mix Recommendations
  • Revenue Forecast
  • Margin Opportunity

All material pricing actions require human approval. No price is ever changed automatically in V1.

Commercial Performance Agent

Measure commercial outcomes and continuously improve AI decisions.

Learning

Reasoning runs today

12

Recommendations

9

Acceptance

100%

Realised value

Reasons across

Recommendation outcomesControl groupsRealised revenueRealised marginAdoptionSeasonality

Produces

  • Fleet Utilisation
  • Revenue per Available Vehicle
  • Revenue per Rental Day
  • Margin
  • Add-on Penetration
  • Upgrade Rate
  • Direct Booking Share
  • Retention
  • CLV
  • Agent Recommendation Performance

Learns which recommendations actually create incremental value and re-weights future agent output.

Shared signal layer

All agents read the same uploaded data and the same knowledge base — no agent has a private view of the truth.

Signals

Bookings, fleet, availability, transactions, rates, movements, maintenance, damage, CSI, channel and margin data from the Data Hub.

Knowledge

Rental, fuel, fine, insurance and upgrade policies, vehicle catalogue, add-ons, branch information, leasing and corporate programmes.

Guardrails

No automatic price changes. Human escalation on complaints, damage and fines. Every recommendation carries confidence and expected impact.