Product recommendations for Ahlsell’s website and app
Personalized product recommendations for Ahlsell’s customers
We helped out with kickstarting the in-house AI capabilities at Ahlsell by collaborating on the development of a recommender system for their website and smartphone app.
Background
With 33 billion SEK in revenue and 5700 employees, Ahlsell is one of the largest suppliers of installation products, tools, and supplies to installers, contractors, facility managers, industry, energy companies, and the public sector in the Nordics.
The business-to-business segment is experiencing the success of digitalization. For Ahlsell it is therefore of equal importance to assist and aid their customers on the web as it is in each physical store. Thus, the recommendations their customers see should be personalized and relevant, so that they can concentrate on their work.
Solution
In close collaboration with Ahlsell’s data scientists, the team developed an end-to-end pipeline handling data ingestion, data processing, and model predictions. The recommendation system consists of a collaborative filtering model as well as a content-based model. A graphical interface was built to evaluate the produced recommendations continuously during the development phase. The pipeline was integrated with the front-end and serves fresh recommendations according to a schedule.
Tools/Tech
We built data ingestion pipelines in close collaboration with Ahlsell’s data engineers using various Microsoft Azure services. Model training and validation pipelines, scheduled re-training, and deployment were set up in Azure Machine learning resulting in a fully automated setup. We wrote the codebase in Python, using various well-tested open-source libraries.
We want our department to be modern, process-light, and with a focus on building products rather than being a general support function, and these drivers were aligned with Modulai’s way of working.
Johanna Staaf, Head of Labs and Applied AI