Modulai helped kickstart in-house AI capabilities at Ahlsell by co-developing a personalized recommender system for their website and smartphone app. The system combines collaborative filtering and content-based models to serve relevant product recommendations to Ahlsell's B2B customers, helping them find what they need faster.
Stats
33B SEK
Annual revenue
5,700
Employees
Challenge
With 33 billion SEK in revenue and 5,700 employees, Ahlsell is one of the largest suppliers of installation products, tools, and supplies in the Nordics, serving installers, contractors, facility managers, industry, energy companies, and the public sector. As the B2B segment increasingly moves online, it is equally important for Ahlsell to assist customers on the web as it is in physical stores. Personalized, relevant recommendations help customers concentrate on their work rather than searching for products.
Solution
In close collaboration with Ahlsell's data scientists, the team developed an end-to-end pipeline handling data ingestion, processing, and model predictions. The recommendation system consists of a collaborative filtering model and a content-based model. A graphical interface was built to continuously evaluate the produced recommendations during development. The pipeline was integrated with the front end and serves fresh recommendations on a schedule.
Tools
Data ingestion pipelines were built in collaboration with Ahlsell's data engineers using various Microsoft Azure services. Model training, validation, scheduled retraining, and deployment were set up in Azure Machine Learning, resulting in a fully automated setup. The codebase was written in Python using well-tested open-source libraries.
Value created
Built alongside Ahlsell's own data scientists, the system gave them both a working recommender in production and the in-house capability to run and develop it further. It serves personalized recommendations across Ahlsell's web and app, helping B2B customers find the right products faster across a catalog built for many different trades.
Learn more
