A Q&A with Ahmed Mhiri, CEO of Free2move
Artificial intelligence has become one of the most talked-about forces in business, and in mobility, its real value is not in the buzz, it is in the execution. At Free2move, AI has been embedded into the company’s platform for years, helping power the real-time decisions required to operate at scale across millions of trips, millions of users and thousands of vehicles.
In this Q&A, Ahmed Mhiri, CEO of Free2move, discusses how the company built its own machine learning capabilities, how AI improves fleet operations and customer experience, and why the next frontier is using AI to deliver more personalized service at mass-market scale.
Q: Everyone is talking about AI. What does AI actually mean inside Free2move?
Ahmed Mhiri: For us, AI is not a trend or a marketing layer. It is an operational capability.
Free2move operates in an environment defined by constant movement: millions of trips, millions of users, thousands of vehicles and countless real-time decisions that affect customer experience, fleet efficiency and financial performance. In that context, AI is not a future concept — it is an operational necessity.
We began building our own machine learning capabilities long before AI became mainstream because we needed to understand and manage mobility at scale. When you process large volumes of structured data and high-frequency events, traditional decision-making methods are no longer enough. Machine learning became the only practical way to identify patterns, anticipate demand and act in real time across complex urban environments.
Q: Can Free2move say what exact AI system, model or partner it uses?
Ahmed Mhiri: The most accurate way to describe our approach is that Free2move has built its own machine learning stack over time, and we continue to enhance it with open-source technologies and internal expertise.
This stack supports number-centric and deterministic use cases such as dynamic pricing, fleet relocation, churn prevention, risk scoring, driver authentication and other real-time decision-making processes.
For customer-interaction use cases, we have tested several large language models. Some currently meet our needs, but the market is evolving quickly, so we continue to monitor and test available solutions. What matters most to us is not the name of a model or provider — it is reliability, scalability and the ability to deliver measurable value for customers and operations.
Q: Was there a specific “aha moment” when Free2move decided to integrate AI, or was it a gradual progression?
Ahmed Mhiri: It was a natural progression driven by the reality of the business.
We had vast amounts of structured data and a very high volume of events across our platform. Very early, it became clear that machine learning was the only way to learn at scale. The “aha moment,” if there was one, came when we looked at demand patterns in a large city.
Urban mobility is incredibly complex. Demand changes by neighborhood, time of day, weather, events, commuting habits and many other variables. To serve customers well, we needed to understand where and when people would need vehicles before that demand appeared.
That led us to build digital models, which we call digital twins, of our cities. Thanks to years of granular demand data, we can now better predict where and when the next user will need a car.
Q: How does that intelligence improve the customer experience?
Ahmed Mhiri: Convenience is everything in mobility. If a customer opens the app and cannot find a vehicle nearby, the service loses value. AI helps us improve that experience by making the system more predictive and responsive.
For example, our demand models help us decide where vehicles should be positioned, when they should be relocated and how to improve utilization across the fleet. That means more customers can find vehicles where and when they need them, and the fleet can be used more efficiently.
The customer may not see the AI directly, but they feel the result: better availability, faster service, fewer friction points and a more reliable experience.
Q: Which areas of the Free2move platform are powered by machine learning today?
Ahmed Mhiri: Our machine learning foundation supports several critical areas of the platform. These include:
- Dynamic pricing
- Fleet relocation
- Churn prevention
- Risk scoring
- Driver authentication
- Real-time operational decision-making
- Demand prediction
- Customer feedback analysis
- Customer support automation
These are not isolated innovation projects. They are embedded into the daily mechanics of how a mobility platform performs.
Q: How are large language models changing Free2move’s approach to customer interaction?
Ahmed Mhiri: Traditional machine learning helps us optimize operations. Large language models open a different opportunity: understanding the voice of the customer at scale.
We can now analyze and react to 100% of customer feedback comments. That allows us to detect trends, identify recurring issues earlier and trigger the right actions faster. In a service business, responsiveness is just as important as availability.
The goal is not simply to automate conversations. The goal is to understand customers better, reduce friction and improve the service continuously.
Q: Can you share a concrete example of AI improving operations?
Ahmed Mhiri: A good example is the end-of-rental process.
It sounds simple: a customer finishes a trip and wants to end the rental. But in practice, several issues can prevent a rental from closing successfully. A user might leave a window open, or park outside the operating area.
Before we introduced an in-app support bot for this journey, these situations represented about 10% of calls into our customer-service teams. By adding conversational support directly in the app to guide customers through the process, we reduced total customer-service calls by about 20%. And 85% of users who interacted with the bot were helped immediately, without escalation.
For the customer, that means less frustration. For the business, it means better service at scale.
Q: How long does it take to implement these AI capabilities?
Ahmed Mhiri: Much of our platform has been built over multiple years of testing, learning and improving. We do not think of AI as something you implement once and finish. It is continuous.
That said, the speed of experimentation has increased significantly. With AI coding assistants, our teams can move faster from idea to test. In one recent example, we took a new critical pricing logic from ideation to reliable production testing in just four weeks.
The important point is that speed must be balanced with discipline. In mobility, real-world reliability matters.
Q: What are the biggest challenges in using AI at Free2move’s scale?
Ahmed Mhiri: Our biggest challenge is also one of our strengths: scale.
When you operate millions of trips for millions of users with thousands of vehicles, your risk tolerance has to be very low. One glitch can create significant financial and operational impact very quickly.
That means we have to test carefully before deploying AI in critical areas. We are excited by the potential of new models, but we do not treat them as magic. In many use cases, 99% accuracy is not enough, especially when customers are asking about pricing, terms and conditions, or operating processes.
Trust has to come before automation.
Q: What has Free2move learned from deploying AI so far?
Ahmed Mhiri: We have learned three major lessons.
First, AI is not plug-and-play. It takes time and effort to test, train and adapt models to perform specific tasks well.
Second, people matter as much as technology. Some people are naturally excited by these tools and others are more cautious. We have seen the strongest progress when we give AI projects to people who spontaneously show interest in the technology.
Third, there is a lot of marketing around AI right now. Many providers are announcing AI capabilities, but very few truly understand how to use AI at scale in a live operational environment. That is where real experience matters.
Q: How does AI connect to Free2move’s broader mobility strategy?
Ahmed Mhiri: Free2move’s mission is to give customers access to the right mobility solution at the right time through one digital ecosystem. AI helps make that possible.
Whether the customer needs car-sharing for a short urban trip, short-term rental for a weekend, or a more flexible longer-term solution, the platform has to understand demand, optimize availability, manage risk and deliver support efficiently.
AI allows us to make that ecosystem smarter, more responsive and more personalized.
Q: What is the next frontier for AI at Free2move?
Ahmed Mhiri: The next frontier is using AI to bridge the gap between scale and personalization.
Mobility businesses have traditionally had to choose between efficient mass-market service and high-touch personalized support. AI creates the possibility of delivering both.
Our ambition is to reduce incidents, raise service levels and bring a more personalized, white-glove account management experience to the B2C mass market. In simple terms, we want customers to feel like the service understands them, anticipates their needs and helps them instantly when something goes wrong.
Q: What does success look like?
Ahmed Mhiri: Success means mobility that feels effortless.
It means fewer incidents, faster resolution, better vehicle availability and a more intuitive customer journey. It means customers spend less time dealing with problems and more time moving.
For Free2move, AI is not about replacing the human dimension of mobility. It is about making mobility more responsive, more reliable and more aligned with real customer needs.
The future of mobility will belong to platforms that can learn faster, adapt faster and serve customers better. AI is helping us build that future.
About Free2move
Free2move is a global mobility provider offering a complete and unique ecosystem to its individual and business customers. Driven by data and technology, Free2move makes the customer experience its top priority. Clean, safe, affordable and accessible via a single app, the offering includes free-floating car-sharing, short, medium and long-term car rental, subscription-based car-sharing and parking services. Free2move currently has more than six million customers, 450,000 rental vehicles and 500,000 parking spaces. Headquartered in Paris, the company is part of the global automotive manufacturer and mobility provider Stellantis.
For further information: https://www.free2move.com
