Fullcargo: commercial vehicle requests resolved by AI, with Karve behind it
Fullcargo's customer does not ask for "a car": they ask for a refrigerated van for Tuesday or a large van with a date change. An AI agent that knows the fleet and the industry, integrated with Karve, now resolves 73% of those sales conversations with no team involved.
Summary
- Fullcargo rents vans and commercial vehicles in the Canary Islands; every request carries load, licence or equipment requirements.
- A generic chatbot could not take it on: it took understanding the request, knowing the fleet and quoting with live availability.
- The agent, integrated with Karve, proposes the right vehicle, handles date changes and recognises customers by their phone number.
- 73% of sales conversations are resolved without the team, and one in ten quotes ends in a confirmed booking.
When people talk about selling vehicle rental by chat, they almost always picture cars: dates, category, price. Commercial vehicles are a different world. The customer knows what they need and describes it in detail: a refrigerated van for a delivery, a high-volume van for a removal, a vehicle that can be driven on a standard licence, often with special conditions and changes along the way.
Fullcargo, in the Canary Islands, is a good test of whether an AI agent can handle that. This case looks at what it took for the conversation to end in a booking written in Karve, and what changed for a team that used to read and check every request by hand.
1. A commercial vehicle rental company in the Canary Islands
Fullcargo rents vans and specialised commercial vehicles to businesses and individuals in the Canary Islands. Its reservation system is Karve, and its fleet has characteristics that matter in every booking: load capacity, refrigeration equipment, dimensions, required licence.
Every enquiry needed someone on the team to understand the case, check which vehicle fitted and was available, and reply. A share of those enquiries arrived outside service hours and waited until the next day.
2. The challenge: specialised requests a generic bot does not understand
The problem was not volume but complexity. A typical request mixes several requirements and tends to change during the conversation: dates move, a new condition appears, the customer asks whether a specific vehicle works for their job. An FAQ chatbot cannot reason about that, and a custom-built one has to be taught the industry from scratch.
Fullcargo needed an agent that already knew how a rental works, that knew its fleet in detail and that quoted against Karve with real availability and prices.
3. What was set up: an agent that knows the fleet and the industry
A car rental AI agent was integrated with Karve and switched on for WhatsApp and the website chat. The difference from a custom bot is the starting point: the agent is already trained in the industry, knows what gets asked and what gets checked before booking, and on top of that Fullcargo's fleet was loaded with its characteristics.
- Interprets the specialised vehicle request, proposes the right one with Karve's live availability and price, and handles date and condition changes in the same conversation.
- Recognises customers registered in Karve by their phone number, retrieves their details and does not ask again for what it already knows.
- Routes questions about ongoing rentals to the company's support channel.
- When it escalates, the team works from the conversation itself, with the request already understood, shortcuts for the usual replies and direct quotes against Karve, a few seconds per request.
4. Results
- 73%
- of sales conversations handled by the AI with no team involved
- 1 in 10
- quotes ends in a confirmed booking
- 35%
- of conversations arrive outside office hours
Specialised vehicle requests that always used to need a person are now resolved in the conversation. For the team, the change is less noise: far fewer requests, alerts only when its judgement is needed, through a notification system that respects its working hours, and every request handled from the conversation in seconds.
“Our customer doesn't ask for a car: they ask for a refrigerated van for Tuesday or a large van with a date change. Bota understands it and quotes it, and only alerts us when we are really needed.”
5. What you can copy from this case
- If your fleet has attributes that decide the booking (load, refrigeration, licence, dimensions), load them into the agent: without them it can only ask, not propose.
- Ask for an agent that already knows vehicle rental; teaching the industry to a custom bot is the hidden cost of many projects.
- Identify customers by phone number against your system: it saves questions and treats them as known customers.
- Take care with how you alert the team: few notifications, within working hours, and the request resolved from the same conversation.
Frequently asked questions
Can an AI agent tell two similar vans apart?
Yes, if it knows the fleet's characteristics. At Fullcargo the agent knows which vehicle is refrigerated, how much it carries or which licence it needs, and proposes the one that fits the request.
What happens when the customer changes dates mid-conversation?
The agent quotes again against Karve with the new dates within the same conversation, without the customer having to start over.
How does the team step in when needed?
It receives an alert with the request already understood and works from the conversation itself, with shortcuts and direct quotes against Karve. Each request takes a few seconds.
Written by
Johan Smith
Car Rental Editorial Team
Sources
- Fullcargo case study on Bota — figures and implementation details published by the customer.
- Fullcargo — the commercial vehicle rental company's website.
- Karve — reservation system used in this case.