Many retailers struggle to provide a personalized shopping experience to their clients. One of the most common approaches retailers can use to rise up to this challenge is through the use of product recommendation systems. Recommendation systems are usually built by making use of direct signals from client feedback. However, many retailers do not have access to this type of information. This case study aims to explore different techniques to create a reliable recommendation system that could be used in a retail business, by making use of only implicit information in lieu of direct client feedback. The recommendations techniques covered throughout the paper consist of sequenced based recommendation systems, recommendation systems through the use of different types of Collaborative Filtering models and hybrid modeling approaches, among others. This case study also covers a high-level view of techniques for handling the issue of not having enough data on a client to make proper recommendations via common modeling techniques. Based on the results from this project the recommendations are to make use of a sequence-based modeling approach using neural networks, along with a popular product recommendation model. Subsequently, the paper aims to present the results in a way that could be both meaningful and actionable in a retail business.