Q&A with Free2move CEO Ahmed Mhiri
Artificial intelligence has become one of the most talked about forces in the business world, but when it comes to mobility, its real value lies in the execution, not the talk. At Free2move, AI has been built into the company’s platform for years, supporting the real-time decision-making required for large-scale operations across millions of trips, millions of users, and thousands of vehicles.
In this Q&A, Free2move CEO Ahmed MiriLearn how the company built its unique machine learning capabilities, how AI improves fleet operations and customer experience, and why The Next Frontier is using AI to deliver more personalized services at mass-market scale.
Q: Everyone is talking about AI. What does AI actually mean within Free2move?
Ahmed Miri: For us, AI is not a trend or a marketing layer. It’s operational ability.
Free2move operates in an environment defined by constant movement: millions of trips, millions of users, thousands of vehicles, and countless real-time decisions that impact customer experience, fleet efficiency, and financial performance. In that context, AI is not a futuristic concept but an operational necessity.
We started building our own machine learning capabilities long before AI became mainstream because we needed to understand and manage mobility at scale. Traditional decision-making methods are no longer sufficient when dealing with large amounts of structured data and high-frequency events. Machine learning has become the only practical way to identify patterns, predict demand, and act in real-time across complex urban environments.
Q: Can Free2move say exactly which AI systems, models, or partners it uses?
Ahmed Miri: The best way to describe our approach is that Free2move has built its own machine learning stack over time and continues to enhance it with open source technology and in-house expertise.
This stack supports number-centric, deterministic use cases such as dynamic pricing, fleet relocation, churn prevention, risk scoring, driver authentication, and other real-time decision-making processes.
For the customer interaction use case, we tested several large-scale language models. While some solutions currently meet our needs, the market is rapidly evolving and we continue to monitor and test available solutions. What matters most to us is not the model or provider name. reliability, scalability, and the ability to deliver measurable value to customers and operations.
Q: When Free2move decided to integrate AI, was there a particular “aha” moment, or was it a gradual process?
Ahmed Miri: It was a natural progression driven by business realities.
Our platform had a huge amount of structured data and a very large amount of events. Very early on, it became clear that machine learning was the only way to learn at scale. If there was ever an “aha” moment, it happened when we observed demand patterns in large cities.
Urban mobility is extremely complex. Demand varies by region, time of day, weather, events, commuting habits, and many other variables. To serve our customers well, we needed to understand when and where people needed a vehicle, before demand existed.
This led us to build digital models of cities (called digital twins). Thanks to years of detailed demand data, we can now better predict when and where the next user will need a car.
Q: How does that intelligence improve the customer experience?
Ahmed Miri: When it comes to mobility, convenience is everything. If a customer opens the app and can’t find a vehicle nearby, the service loses value. AI helps improve the experience by making systems more predictive and responsive.
For example, our demand model helps you decide where to deploy vehicles, when to redeploy them, and how to improve overall fleet utilization. This means more customers can find their vehicles when and where they need them, and use their vehicles more efficiently.
Customers may not be able to see AI directly, but they are feeling the results: increased availability, faster service, fewer points of friction, and more reliable experiences.
Q: What areas of the Free2move platform are currently powered by machine learning?
Ahmed Miri: Our machine learning foundation supports several key areas of the platform. These include:
- dynamic pricing
- fleet relocation
- Churn prevention
- risk scoring
- driver certification
- Real-time operational decision making
- Demand forecast
- Customer feedback analysis
- Customer support automation
These are not isolated innovation projects. These are built into the day-to-day fabric of how mobility platforms work.
Q: How does a large-scale language model change Free2move’s approach to customer interactions?
Ahmed Miri: Traditional machine learning helps optimize operations. Large-scale language models open up another opportunity: understanding the voice of the customer at scale.
You can now analyze and act on 100% of your customer feedback comments. This allows you to detect trends, identify recurring issues early, and trigger appropriate actions faster. In a service business, responsiveness is as important as availability.
The goal is not just to automate conversations. The goal is to better understand your customers, reduce friction, and continually improve your service.
Q: Please give me a specific example of business improvement using AI.
Ahmed Miri: A good example is the end-of-rent process.
It seems simple. The customer is done with the trip and wants to end the rental. However, in reality, several issues may prevent your rental from ending successfully. Users may leave windows open or park outside the operating area.
Before this initiative introduced in-app support bots, these situations accounted for approximately 10% of calls to the customer service team. By adding conversational support directly into the app to guide customers through the process, we reduced the total number of customer service calls by approximately 20%. Additionally, 85% of users who interacted with the bot were helped immediately without escalation.
For customers, that means less frustration. For businesses, that means improved service at scale.
Q: How long will it take to implement these AI features?
Ahmed Miri: Many of our platforms have been built over years of testing, learning, and improvement. We do not believe that AI can be implemented once and then be over. It’s continuous.
However, the speed of the experiment has been greatly improved. AI coding assistants help teams move from idea to test faster. In one recent example, we took critical new pricing logic from ideation to reliable production testing in just four weeks.
The important point is that there must be a balance between speed and discipline. In mobility, real-world reliability is important.
Q: What are the biggest challenges in using AI at Free2move’s scale?
Ahmed Miri: Our biggest challenge is scale, which is one of our strengths.
When you operate millions of trips for millions of users using thousands of vehicles, your risk tolerance must be very low. A single issue can quickly have a significant financial and operational impact.
This means that AI must be carefully tested before deploying it in critical areas. We’re excited about the potential of the new model, but we don’t consider it magic. For many use cases, 99% accuracy is not enough, especially when customers ask questions about pricing, terms and conditions, or operational processes.
Trust must come before automation.
Q: What has Free2move learned from implementing AI so far?
Ahmed Miri: We learned three big lessons.
First, AI is not plug-and-play. Testing, training, and adapting models to perform specific tasks well takes time and effort.
Second, people are just as important as technology. Some people are naturally excited about these tools, while others are more cautious. The greatest progress is made when AI projects are given to people who have spontaneously expressed an interest in the technology.
Third, there is currently a lot of marketing related to AI. Many providers are announcing AI capabilities, but few truly understand how to use AI at scale in real-world production environments. That’s where real experience becomes important.
Q: How does AI fit into Free2move’s broader mobility strategy?
Ahmed Miri: Free2move’s mission is to give customers access to the right mobility solution at the right time through one digital ecosystem. AI can help make that possible.
Whether a customer requires car sharing for a short urban trip, a short-term rental for the weekend, or a more flexible long-term solution, the platform must understand demand, optimize availability, manage risk, and provide support efficiently.
AI allows that ecosystem to become smarter, more responsive, and more personalized.
Q: What is the next frontier for AI at Free2move?
Ahmed Miri: 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 services and highly personalized support. AI creates the possibility to do both.
Our goal is to reduce incidents, improve service levels, and provide a better and more personalized account management experience to the B2C mass market. Simply put, you want your customers to feel like your service understands them, anticipates their needs, and helps them right away when something goes wrong.
Q: What does success look like?
Ahmed Miri: Success means mobility that feels effortless.
This means fewer incidents, faster resolution, increased vehicle availability, and a more intuitive customer journey. This means customers spend less time dealing with issues and more time traveling.
For Free2move, AI does not replace aspects of human mobility. It’s about making mobility more responsive, reliable, and tailored to real customer needs.
The future of mobility belongs to platforms that can learn faster, adapt faster, and better serve customers. AI can help us build that future.
About Free2move
Free2move is a global mobility provider that offers a complete and unique ecosystem for residential and corporate customers. Free2move is data and technology driven and puts customer experience first. Clean, safe, affordable and accessible through one app, the service includes free floating car sharing, short-term, medium-term and long-term car rentals, and subscription-based car sharing and parking services. Free2move currently has over 6 million customers, 450,000 rental cars and 500,000 parking spaces. Headquartered in Paris, the company is part of Stellantis, a global car manufacturer and mobility provider.
For more information: https://www.free2move.com
