Entity Matching of Product List from 2 Retailers

We are in the business of identifying matches of retail products between two different product lists based on Product Descriptions on different e-commerce sites. We currently use fuzzy lookup in Excel to instantly identify high likelihood exact matches, then we start the laborious process of validating proposed matched pairs with lower match scores. In this manual process, we have created several datasets that have identified False Positives & True Positives (False Negatives), which can be used in training datasets. In this manual process, we also utilise other information such as matches by Brand Name, Categories, Product Weight/Volume (range), or Price (range) to either validate the fuzzy match or identify the correct match; however, we find that all this work is not being invested in training and refining a model to get better subsequent matches. We do not have programmers or coders who can help us with any machine learning modules or libraries. For this project, we are looking for someone who can use code libraries from entity matching models such as Magellan or work with 3rd party entity matching API (e.g. Azure ML or Auto EM) to create a learning program that can propose better and better matches each time. This person will also help us train the model by taking our validation, matching methodology, and training datasets and re-run the matching on the improved model. Once, we're happy with the model, this person will also help us create an interface or program (e.g. web app or Excel) where we can conduct validation of matches AND identify False & True Positives that will be used to train the model even further. Finally, this program will need to store the model and user interface in our Azure account AND export the output in CSV or Excel worksheets.