Artificial intelligence: Initial case studies

Cas pratiques IA
February 16th, 2021
Share :
{{totalComments}} comments

At the Universités des Achats 2020 event, one workshop raised the topic of artificial intelligence and highlighted some of the ways it could help procurement departments.

Four speakers covered the subject:

  • Pascal Goumand, Head of Purchasing, Savings, Insurance and Property at Crédit Agricole
  • Thomas Solignac, CEO and Co-Founder of Golem.ai, a start-up specialising in Artificial Intelligence (AI) solutions 
  • Julie Dang Trang, Managing Director of Manutan France
  • Finally, Xavier Laurent, Mergers and Acquisitions Director for the Manutan Group

When can we turn to artificial intelligence?

First and foremost, we need to identify the processes to be optimised with new technologies. A McKinsey study estimates that the vast majority of the procure-to-pay process, i.e. from searching for products to paying the invoice, can be automated. Some of the most easily automated processes include order and invoice reconciliation, order monitoring and reporting, and invoice management.

Beyond these theoretical figures, Pascal Goumand suggests identifying the most time-consuming tasks with no real added value for your procurement team in order to prioritise investment. For example, at Crédit Agricole, this includes contract negotiation, expenditure analysis and ensuring compliance with new regulations (the French Sapin II anti-corruption law, duty of care, GDPR etc.).

Finally, artificial intelligence will make it quicker and easier to accomplish these tasks, so teams can focus instead on assignments with higher added value. These new assignments could take many forms, for example helping other departments review their needs, improving skills across the team and monitoring the markets. Pascal Goumand's last piece of advice for deciding what AI investments to make first is to think about ROI (Return on Investment) in the short and long term.

Artificial intelligence case studies

During the workshop, the participants discussed a variety of case studies relating to different fields. Two of them are particularly relevant and inspiring for procurement.

Contract management at Crédit Agricole

This project, which is still in the proof-of-concept phase, is aiming to accelerate and improve the quality of contract management for a variety of reasons. Pascal Goumand uses the example of GDPR, which meant that all the contracts for the entire Group had to be analysed to ensure their compliance. At the time, it took the procurement team almost six months to identify all the contracts that contained personal data and verify that these contracts also included the required clauses.

The company is now testing an AI solution that will enable this work to be done automatically. As a result, the procurement department estimates that processing time will be six times faster. The department will soon be trialling AI for constant contract monitoring.

Product classification at Manutan

To optimise the classification of product data from its suppliers, Manutan has rolled out an artificial-intelligence solution with the help of start-up Factonics.

This task is now fully automated, and teams need only validate some pre-classified products that have not achieved a certain accuracy score. Manutan anticipates being able to save 1 to 2 weeks per new listing file, which equates to almost 280 days a year!

Thanks to these concrete examples, the benefits in terms of time and productivity are clear. If you want to take the plunge, we recommend watching the whole video to learn more about the best practices for selecting the right start-up for you and working with them in the best conditions.

Livre blanc
white paper
Optimise your long tail spend with our white paper on e-procurement