I’m looking to build a solution where we predict which product best matches a customers requests for quotation.
An example would be that a customer requested “Heavy Duty Glass beaker 150ml” and the correct match would be BEAKER - SUPER DUTY - LOW FORM - DURAN 150ml
We have recently implemented a workflow where we are capturing the customers request and the match product, which we can then use as training data.
A possible solution where more than one product is returned with a probability score is one possible scenario.
Does anyone have any advise on how to tackle such a project.