At UNiDAYS we work with many brands who provide us with product level data for transactions. In this talk we will walk through the end to end steps we took at UNiDAYS to build a system to process this data and turn the text data into a product catalogue, using NLP/NER. We will also look at many of the problems which arose during the building of the system and how we overcame them. additionally we will talk through where we decided to use AI and when we chose not to and the factors which lead us to these decisions.
Technical level: Technical practitioner
Session Length: 40 minutes
Supported by
