La Poste

La Poste radically accelerates its AI projects with Prevision Enterprise AI Platform. has enabled us to reduce the time-to-market of AI projects by a factor of 4.

Rachid Alili,

Chief Data Scientist & Data Governance / Groupe La Poste

About La Poste

La Poste is one of France’s largest companies and is responsible for its postal services. Today, La Poste is a diversified group with significant retail banking, logistics, parcel delivery, and digital services accounting for a growing part of its revenue. Groupe La Poste is present in 44 countries across the globe and generates 30 percent of its 26 billion euro revenue internationally.

A quest to optimize the last mile

Last-mile parcel delivery has become a critical, and intensely competitive, area of the digital economy.

Letter and parcel operations have a lot in common. However, unlike letters, the recipient of a parcel typically pays for the service on top of a purchase. For this reason, online shoppers are extremely demanding in terms of speed, convenience, reliability and visibility. However, they are price sensitive in equal measure. 

Likewise, for retailers, last-mile logistics is a key component of the economic equation that typically represents a substantial chunk of shipping costs. It is also a make-or-break factor in the customer experience.

With pressure to optimize delivery routes and processes, merchants and logistics players all practice a form of arbitration: they subcontract to the lowest bidder.

Using enterprise AI better to predict delivery slots

La Poste wanted to increase the visibility and control it had over the timing of its deliveries, which themselves often depend on sub-contractors. In effect, at the time, the precision of its estimated delivery slots was limited to half a day or 2 hours at best. However, these estimates were typically inaccurate. One direct consequence of which was that the stated delivery slots given to customers were not respected.

La Poste had track-and-trace data at its disposal. However, it was not telling the full story as operators were scanning articles in batches. This data was difficult to exploit because multiple articles were sharing the same timestamp and location data and were therefore more difficult to reconcile.

In order to both improve customer satisfaction and gain an operational edge over the competition, La Poste decided to use artificial intelligence to improve the reliability of its delivery slots.

Using the Prevision platform, La Poste’s data science teams were quickly able to harness the disparate tracking data provided from its own operations and those of its business partners. This allowed La Poste to model the behavior of actual delivery times. 

The resulting data app is able to more reliably predict delivery slots than alternative solutions. La Poste is benefiting from enhanced operational visibility. Most notably, shoppers now receive advance notification of 1-hour delivery slots, which has led to a substantial improvement in customer satisfaction and much lower volumes of complaints.

Accelerated data science for all

La Poste now plans to use the Prevision platform to analyze international parcel paperwork. This will enable it to automatically categorize items and to apply the correct customs duty in the face of growing volumes — even on lower value items, as is increasingly required by the regulator. 

This initiative is part of a broader move to empower La Poste’s business users to become citizen data scientists. In other words, to be more autonomous in their ongoing data science projects and thus free up the resources of the central data science teams.

In effect, faster cycles are essential not only to increase agility and to reduce costs but also to enable La Poste to reduce the opportunity cost of exploring proof-of-concept projects.

We want to democratize AI, to make it more accessible and explainable to our business users, and deliver on its promises faster,” said Rachid Alili, Chief Data Scientist at Groupe La Poste. “We’ve immediately benefited from a 10-fold improvement in infrastructure and IT run costs compared to our targets. Importantly, the platform has enabled us to reduce the time-to-market of our AI projects by a factor of 4.”

Time to market reduced by a factor of 4

More effective use of data scientist resources

Improvement in infrastructure and IT run

Accelerate Your Enterprise AI Projects Today

Create a free trial account to use the platform now. No credit card required.
Or get in touch to discuss your business challenges or schedule a custom demo.

Get started