Overview
Freigent uses specialized AI agents that each handle a different part of carrier operations. They're not chatbots — they're workers that take action. They read emails, make decisions, send responses, generate documents, and flag issues.
How AI Autonomy Works
You control how much the AI does on its own. There are four levels:
Manual Mode
The AI watches but doesn't act. It shows you recommendations and suggestions, but you make every decision and take every action. Good for when you're learning the system or for high-stakes decisions.
AI Suggest Mode
The AI analyzes the situation, drafts a response or action, and presents it to you. You review, edit if needed, and click "send" or "approve." The AI does the thinking, you do the clicking.
AI Approve Mode
The AI handles routine tasks on its own but pauses for your confirmation on important decisions. For example, it might auto-send a status update but wait for your OK before accepting a rate below your target.
AI Auto Mode
The AI handles everything within the rules you've set. You review results after the fact. The AI will still escalate truly unusual situations — a rate that's way outside market norms, a compliance violation, a broker dispute.
You can set different autonomy levels for different departments. Maybe you run full auto on status updates and invoicing, but keep rate negotiation on AI Approve.
The Agents
Dispatch & Booking Agent
What it does: Handles the entire load lifecycle from offer to booking.
- Reads incoming emails from brokers and extracts load details
- Evaluates rates against market data and your preferences
- Negotiates via email (counter-offers, rate justifications)
- Books loads that meet your criteria
- Sends rate confirmations
- Assigns drivers based on the ranking engine
Example: A broker emails offering Dallas to Houston, dry van, $2,100. The AI checks: market rate is $2,350, this broker has a 75% reliability score, you have a driver 45 miles from pickup with 9 hours available. AI counters at $2,300 citing market rates and your available capacity. Broker comes back at $2,200. AI accepts (within your rules) and books the load.
Communication Agent
What it does: Handles status updates and routine communication.
- Sends automatic pickup and delivery confirmations
- Provides check-call responses to brokers
- Notifies customers about delays proactively
- Handles routine email inquiries
- Sends driver instructions via the mobile app
Example: A broker sends "Can I get a status update on load #4521?" The AI checks the driver's GPS, sees they're 2 hours from delivery, and responds: "Driver is currently 120 miles out, ETA 2:15 PM. On schedule for the 4:00 PM delivery window."
Compliance & Safety Agent
What it does: Monitors regulatory requirements and flags issues.
- Tracks Hours of Service for all drivers
- Alerts before drivers run out of drive time
- Monitors document expirations (CDL, medical card, insurance, DOT authority)
- Tracks vehicle inspection schedules
- Monitors Drug & Alcohol Clearinghouse status
- Flags compliance gaps in driver qualification files
Example: Driver's medical card expires in 30 days. AI creates a high-priority task, sends the driver a reminder via the app, and blocks new load assignments after expiration if not renewed.
Accounting Agent
What it does: Handles invoicing, payment tracking, and collections.
- Generates invoices with all required documents attached
- Sends invoices to the correct contact (billing email, not dispatch email)
- Follows up on unpaid invoices with escalating urgency
- Tracks payment aging (30/60/90 days)
- Calculates driver settlements
- Handles factoring submissions
- Syncs with QuickBooks or Xero
Example: Load delivers at 2 PM. By 3 PM, the AI has generated the invoice with BOL, POD, and rate confirmation attached, and emailed it to the broker's billing department. At 30 days, if unpaid, it sends a polite follow-up. At 45 days, firmer. At 60 days, it flags for escalation.
Load Monitor Agent
What it does: Watches loads in transit and handles issues.
- Tracks driver location via ELD integration
- Predicts arrival times based on real-time data
- Detects delays and notifies affected parties
- Handles detention tracking (time waiting at pickup/delivery)
- Monitors for route deviations
Example: Driver hits unexpected traffic on I-35. AI detects the delay from GPS data, recalculates ETA, and proactively emails the broker: "Update on load #4521: traffic delay on I-35 near Waco. Revised ETA is 4:45 PM, 45 minutes later than original. Driver is in communication and will update on arrival."
Maintenance Agent
What it does: Tracks vehicle maintenance and predicts issues.
- Monitors vehicle diagnostics from ELD/telematics
- Schedules preventive maintenance based on mileage and time intervals
- Tracks repair history per vehicle
- Alerts on diagnostic trouble codes
- Manages DVIR (Driver Vehicle Inspection Report) workflows
AI Decision Transparency
Every AI decision is logged and visible:
- What the AI decided
- Why (the factors it considered)
- What data it used (market rates, broker history, driver availability)
- The confidence level
- What alternatives it considered
You can always override an AI decision. The system learns from overrides to improve future recommendations.
Setting Up AI Agents
When you first set up Freigent, the AI starts in Manual Mode for everything. As you use the system and the AI learns your preferences, you can gradually increase autonomy:
- Start with AI Suggest for status updates (low risk)
- Move to AI Approve for invoicing (medium risk, clear rules)
- Try AI Approve for rate negotiation (higher value, you still confirm)
- Eventually AI Auto for everything routine
There's no pressure to go full auto. Many carriers run in AI Approve mode for years and are happy — the AI does 90% of the work, they just click approve.