We're back with an exciting Meetup on recommender systems and deep learning this month (July 25th) to keep you cool this summer!
For online businesses, customer satisfaction is key to staying ahead of competition. To provide customers with the best possible experience, companies must shift their outbound marketing strategy from static segmentation to highly personalised recommendations, based on personal unique patterns.
Though it’s common for online businesses to give recommendations for other services or products that consumers might be interested in based on their behavior, high-quality recommendations are extremely difficult to provide in practice. They can consume extraordinary amounts of internal
resources, as they require heavy data manipulation, coding, and testing of different algorithms. So when a data team is tasked with creating a recommender system based off a variety of signals, the project can
get complicated quickly.
Check out our talks below:
Talk #1: Destination Deployment: Gousto's Recipe Recommendation System
Gousto helps make home-cooking easy while minimising food waste, aiming to be responsible for a total of 400 million meals in UK homes by 2025. A big part of enabling our growth is understanding what our customers like to cook and how they plan their meals. Last year, our data science team built and deployed Gousto's first recommendation engine and have since then gone through multiple iterations. In his talk, Marc will go through the team's journey to deploying a graph-based recommendation system using Docker, Airflow and AWS Batch.
Speaker Bio: Marc Jansen
Marc is a data scientist at the UK's leading recipe kit company, Gousto. Since joining in early 2017, he has worked on development and production deployment of algorithms for personalization, stock management and warehouse optimization. Prior to joining Gousto, he completed a PhD in Management Science at the University of Cambridge and spent time as a visiting researcher at the MIT-SUTD International Design Centre in Singapore.
Talk #2: How to Improve your Recommender System with Deep Learning
In this talk, we will describe how Dataiku improved an e-business vacation retailer recommender system using the content of images. We’ll explain how to leverage open datasets and pre-trained deep learning models to derive user preference information. This transfer learning approach enables companies to use state-of-the-art machine learning methods without having deep learning expertise.
Speaker Bio: Alexandre Hubert
Alex is a Lead Data Scientist at Dataiku.
As always there will pizzas and many chilled beers for all!!
See you there!