Machine Learning & Data Engineer

We run the website ---------- and YouTube channel ---------- have over 260 partner companies that all have Data Feeds of the products they sell on their individual websites.

We have over 1.5 Million products that need to be fed into a machine learning program and have them come out in our desired categories on our website so our users can search the products to find deals.

Each data feed is a little bit different. They are CSV Data, noisy data. We have 8 main categories with 4 pending post map. 76 sub categories and 32 sub-sub categories to sort the data into.

We have access to title of the product, description and images.

we last used Bert-based-cased and have also used a distillert model, both haven't been able to accomplish what we need (based on how they are currently built) which is why are posting this job.