Chronopost: Generative AI Revolutionizes Package Management

Chronopost Revolutionizes Logistics with Generative AI

Chronopost, the leader in express delivery in France, recently unveiled two major innovations based on generative artificial intelligence, already integrated into its network. These tools aim to significantly reduce parcel loss and transform customer relations. The first, Parcel Matching, is designed to identify parcels without labels, while the second, the autonomous agent Léonard, efficiently handles customer inquiries.

In the express logistics sector, unlabeled parcels and impatient customers represent constant challenges. Chronopost, a subsidiary of the La Poste group, decided to showcase these two innovative solutions at the VivaTech event. These internally developed tools are already operational on a large scale and promise to significantly enhance service efficiency.

Parcel Matching: A Solution for Unlabeled Parcels

Every week, Chronopost must process hundreds of parcels arriving at its hubs without readable labels. These often anonymous parcels pose a real puzzle for delivery services. The Parcel Matching solution, developed by the internal Robot Factory and launched in November 2025, addresses this issue through a three-step analysis.

The process begins with sorting performed by machine learning, which classifies parcels based on their weight and dimensions. Next, generative AI takes over to visually analyze photographs of the parcels taken from six different angles. This analysis allows for a matching score to be generated in just two to four seconds. When this score exceeds 8/10, a human expert verifies the match before reintegrating the parcel into the delivery circuit. Thanks to this method, Chronopost manages to locate approximately 300 parcels each week without requiring a claim.

Benoît Frette, president of Chronopost, stated that this technology allows for the identification of an unlabeled parcel in less than four seconds, directly improving the customer experience by reducing delays and errors.

Léonard: The AI Agent Transforming Customer Service

Chronopost has also introduced Léonard, an autonomous agent that goes far beyond simple chatbots. Capable of understanding natural language, Léonard can recalculate delivery times, process requests with operational sites, and initiate investigations in case of anomalies. In the testing phase, Léonard handled over 100,000 calls, achieving an 85% resolution rate for end-to-end requests. This efficiency has led to a 40% reduction in request processing times while maintaining a customer satisfaction level comparable to that of a human advisor.

Currently, Léonard is used for Shop2Shop parcels, a Chronopost service that allows for the delivery of a package from one relay point to another. Starting in July 2026, this agent will be extended to all of Chronopost's services. This extension is necessitated by a 32% increase in customer service calls over the past five years, primarily due to the growth of e-commerce and shipments between individuals.

Benoît Frette emphasized that this technology frees advisors from repetitive tasks, allowing them to focus on more complex cases. By placing humans at the center of customer service, Chronopost hopes to improve service quality while optimizing its operations.