Examples

We are passionate about helping you capitalize on your data content in this ever-evolving digital world. With our expertise and today’s tools, you can unlock the true potential of your data and gain a competitive edge.

Wareshop Consulting is led by Fred Ware Ph.D., a machine learning and computer vision expert. He has over 20 years of experience in research and development of production applications & systems, dataset curation, data analytics, core processing pipelines, and technology demonstrations.

What We Stand For

Let us show some real examples how to machine learning can be quickly employed on any type of data.

Inventory Management

The Daily Operation: You have inventory and personal supplies that regularly come into your office. Your inventory is regularly shipped out to customers, and your supplies are regularly consumed due to work activities. However your expenses are not added up and their have been inventory disconnects with your customers.

The Goal: You want to keep track of your supplies that you store and ship to another location. Here we showcase how machine learning (object detection) and computer vision (feature matching) can be used to track and alert whenever there are abnormal changes in the supplies you are managing.

The Results: This program took one day to complete and will run on any Windows or Mac OS system. It is using a pre-trained model and general feature matching. Imagine how it can work when it train and configured on your specific products!

Data Analysis of Customer Concerns

The Daily Operation:

You operate an auto shop. Over the years, your customers have voiced their concerns by text messages, emails, and phone calls. Your mechanics and technicians also provide their repair notes in the same reports with the associated customer concerns.

The Goal: You want to objectively understand the sentiments of your customers so you can serve them better. Also you want your mechanics and technicians to know that their concerns are also being heard to keep employee moral up and maintain an enjoyable work environment.

The Results: The first step is to analyze the textual data from your customers and the mechanics. All text message, emails, and phone calls (which are converted to text), are collected into one common dataset. Here we show Word Clouds plots of the “Customer Concerns” and the “Recommended Repair”. Word clouds provide a visual of the frequency of each word in the text data. Most frequently used words are bigger in size. This objective information provides a reference to begin analyzing sentences and generate a ML models that predicts the sentiment of the customers. For example, the model can be trained to proactively inform you what makes the customers happy, satisfied, confused, or frustrated.

The Issue:

Suppose you operate several basketball camps and leagues, and you want to evaluate each participant’s ability to shoot and make jumpers from different areas of the court. This information will allow you properly instruct and development each player according to them specific needs.

The Goal: At every camp event and team practice, you set up one half court for each participant to go and take jump shots according to a design drill.

The Results: Here is a video clip of automatic basketball shot tracker. It utilizes machine learning shot classification and computer vision position estimation. Note: This is an example used a smart phone cameras on a 1 foot tripod. This same approach can be applied to any type of sport.