What if artificial intelligence could do more than answer questions — and support customers throughout their journey? That is the goal of askrnlt, an AI agent designed to deliver a smoother, more personalised and more efficient customer experience. Let’s find more about the first months of agentic AI at Renault with Anthony Vouillon, Director of the Augmented Renault AI Centre of Excellence, and Arnaud Belloni, VP, Renault Brand Global Marketing.
Key takeaways
With askrnlt, Renault is bringing agentic AI into customers’ everyday lives, opening up a new way to support, advise and guide users on renault.fr and renault.es.
Designed as a true digital advisor, this conversational agent supports each visitor in a simple, natural and personalised way throughout their automotive project.
Already delivering promising results, askrnlt is helping to transform Renault’s digital customer experience — a key pillar of our strategic plan.
A real paradigm shift is now taking place, moving from isolated initiatives to a structured approach in which AI becomes a central lever for transforming processes. Available on renault.fr since November 2025 and on renault.es since April 2026, the askrnlt orchestration agent provides a particularly tangible example of this shift. This conversational agent is helping to transform the experience of retail customers, a strategic priority within our strategic plan.
“The question now is not where AI can be used but how AI can fundamentally transform existing processes, such as reinventing the customer experience, and ensure customers have access to consistent and reliable information throughout their journey, from discovery to servicing, including the use of their vehicle.”
askrnlt: the first customer-facing application of agentic AI
Integrated into Renault brand websites, askrnlt is a conversational agent developed in-house by Renault Group’s AI, IS/IT and Renault brand Marketing teams. It guides visitors in natural language, autonomously directing them towards the information and vehicles best suited to their needs.
askrnlt is far more than a technological innovation: it marks a turning point in the way AI is used across the Group. It is the first customer-facing application of agentic AI.
Until now, AI systems have primarily been designed to assist — answering questions, generating content and suggesting recommendations. With agentic AI, the Renault brand is entering a new dimension: specialised agents are capable of carrying out actions autonomously, under human supervision.
In practical terms, askrnlt answers visitors’ questions and offers a simple, personalised and intuitive way to discover Renault vehicles. Through conversation, it helps customers find, compare and choose the vehicle best suited to their project and needs. In France and Spain, the digital experience on Renault brand websites is therefore being reshaped in depth:
A personalised experience: askrnlt adapts its responses based on the conversation, browsing behaviour and Renault brand information to provide more tailored support.
A natural conversation: customers simply speak with askrnlt as they would with an advisor.
Rich responses, illustrated with images, videos and links to Renault website pages, help users move their project forward. The links suggested by askrnlt also reflect what the user has shared during the conversation: for example, if they are looking for a black Renault Austral full hybrid with a roof box, the configurator page to which askrnlt directs them already includes this information, making the journey faster and smoother.
“We designed askrnlt as a support tool for our customers. Whether they are just beginning their thinking or ready to make a choice, askrnlt supports them like an advisor — or a supercharged search engine. It helps guide them towards the option that suits them best. It can anticipate their needs, delivering instant, personalised answers while saving them valuable research time.”
askrnlt features
Conversational agent available on Renault’s website in France since November 2025 and in Spain since April 2026. It now covers the full passenger car range.
Multi-agent architecture with an orchestrator, powered by Gemini 2.5 Flash to ensure fluid, natural interactions.
Data sources: Renault brand website, Renault brand configurator, pricing API, askrnlt confidential database, and Gemini’s general knowledge.
Guardrails put in place by the Augmented Renault programme teams to protect against jailbreak attempts and misuse.
Behind its intuitive interface, askrnlt relies on an advanced architecture. The orchestrator agent analyses each request and activates specialised agents (configurator, product knowledge and more). Each AI agent is connected to multiple data sources (configurator, pricing, product content, etc.). “With askrnlt, our Group is making 2026 the year of AI agent orchestration. As in an orchestra, each agent plays a specific role, coordinated by a ‘conductor’. askrnlt fulfils that orchestration role, with the ability to act under our own supervision.” says Anthony Vouillon, Director, Augmented Renault AI Centre of Excellence
For example:
A customer wants to configure a vehicle → askrnlt, as the orchestrating agent, calls the configurator agent.
They ask a question about a technology—for example how to use the Renault 5’s bidirectional charging → askrnlt mobilizes a knowledge agent.
While the orchestrating agent relies on artificial intelligence, human oversight remains in place to ensure that askrnlt operates as intended: responding appropriately to requests and retrieving the right information from the right sources.
This multi-agent architecture is built on foundations recognised by Mistral AI, a leading European player in generative and agentic AI, particularly for industrial applications. Following its audit, Mistral AI assessed askrnlt as a “well-engineered” solution, citing effective handling of edge cases, well-designed guardrails, and clean code. These conclusions validate askrnlt’s technical credibility and the strength of its foundations for future developments.
AI agents enriching the relationship before the dealership
askrnlt does not replace Renault’s distribution network—it strengthens it.
The goal of askrnlt is to enhance customers’ knowledge upstream, enabling richer conversations with dealers. Put simply, better-informed customers arrive better prepared, their journey becomes smoother, and conversion to purchase is easier. askrnlt is therefore a first building block in a broader transformation of the customer experience within Renault Group — a major focus of the futuREady strategic plan.
“askrnlt is far more than a new feature: it is a game changer. We are reinventing the digital relationship with our customers by making the journey simple, fluid and intelligent. This project reflects our ambition: to make Renault a technology brand that creates desire and delivers unprecedented experiences, thanks to the expertise of our teams and AI developed for our customers.”
By answering customers’ questions throughout a simple, fluid journey, askrnlt puts Renault brand expertise at their service — particularly in the shift to electric vehicles. The AI agent helps them explore the ecosystem and services available across Renault’s electric and hybrid range. askrnlt also helps reassure users. It can explain the ease of using an electric vehicle, clarify charging capabilities, outline the services offered by the Group (charging on the move with Charge Pass, home charging with charging stations, bidirectional charging, etc.) and point users towards website tools such as the route planner, range simulator and charging simulator.
This technological innovation therefore supports one of Renault brand’s missions as the leader in electrified vehicle sales in France: to guide its customers as effectively as possible towards sustainable mobility.
“If there’s one line I’ll stand by: agentic AI only becomes real when it is connected to tools, data, and actions, and when it’s built into an experience people actually use. The goal with askrnlt is not to do full online sales, but rather to direct potential buyers to our dealers, having enriched all their customer knowledge before they even meet the salesperson.”
The askrnlt AI agent, a collective innovation serving the experience
Developed by Group teams, askrnlt is the result of an unprecedented collaborative approach. Augmented Renault AI, Renault Group IS/IT and Renault Marketing teams delivered the project as one team, bringing together complementary perspectives. Strong synergies emerged as the teams worked in an agile way to rapidly design a tool that met business expectations.
This collaborative approach ensured the service’s technical robustness, its seamless systems integration, and above all the quality and relevance of the user experience. For example, the Marketing teams paid particular attention to the customer journey, ensuring that the interaction was not merely informative but genuinely useful at every stage. Strong commitment from the Marketing division helped drive this collective success.
The results already confirm askrnlt’s impact on customer decision-making. Launched in France at the end of 2025, askrnlt is already delivering strong results:
The open rate of askrnlt has tripled compared with the previous chatbot.
On mobile, complete configurations are multiplied by 3 between user without askrnlt and user who had a chat with askrnlt.
60% of the questions relate to vehicles and their configuration first. Financing is the third topic of interest.
Adoption of askrnlt keeps growing, with more than 50 000 chats with askrnlt since its launch.
The arrival of agentic AI on Renault brand websites is therefore contributing fully to a distinctive customer experience. Visitors interact more with the agent and are more likely to complete their configuration journey.
And this is only the beginning. askrnlt will also be rolled out in Italy, Germany, the United Kingdom and Brazil by the end of 2026. Built on a multi-agent architecture, it will continue to evolve with new services and functional building blocks that remain part of a coherent experience — for example, a “stock agent” providing information on vehicles available in stock.
With askrnlt, Renault Group is not simply improving a chatbot for the Renault brand: it is bringing agentic AI into the reality of customer interactions. “This first application of agentic AI lays the foundations for a new model for the Group: AI connected to tools and data; AI capable of acting, not just responding; and AI embedded in an experience that is genuinely useful.” concludes Anthony Vouillon, Director, Augmented Renault AI Centre of Excellence.
Glossary
Generative AI: artificial intelligence capable of producing original content (text, images, videos, audio and/or software code) in response to a prompt or a user request. It relies on models based on the principle of deep learning, which identify and encode patterns and relationships in the enormous quantities of data needed for training, then use this information to understand users’ natural-language requests or questions. These models can then generate text, images and other high-quality content in real time based on the data on which they were trained.
Agentic AI: an artificial intelligence system capable of carrying out tasks with limited supervision. It is made up of AI agents – software entities capable of making decisions in a dynamic context and triggering the actions their AI judges relevant to perform. The term “agentic” refers to the agency (capacity to act) of these models.
Guardrails (or “safeguards”): the set of rules, controls and mechanisms put in place to govern the behaviour of an AI agent. They help ensure that the AI acts safely, ethically and in line with set objectives, by limiting inappropriate actions or responses. In concrete terms, guardrails are used to prevent errors, drift or misuse, particularly when faced with circumvention attempts such as jailbreaks.
Jailbreak: a technique aimed at bypassing the rules, safeguards or limitations imposed on an AI system. It often involves manipulating the instruction (prompt) to lead the agent to produce answers or carry out actions it should normally refuse. In practice, this is used to test an AI’s robustness, but it can also expose security or governance flaws if these protections are insufficient.
Augmented Renault: the Augmented Renault programme aims to structure and accelerate the development of artificial intelligence within Renault Group. It provides teams with the human, technical and tool resources needed to design, deploy and industrialize AI solutions at scale across the company, in connection with IS/IT and business functions. It also relies on an ecosystem of industrial partners, start-ups and research laboratories. Its objective is to transform functions, improve overall performance and offer smarter, more customer-centric products and services.
