SPR

SPR collaboratively worked with the School of the Art Institute of Chicago (SAIC) to deliver a machine learning model that the SAIC admissions office used – and can continue to use – to better predict which students that were admitted will actually enroll at the school. In turn, by better predicting what students will enroll, they can provide those applicants with access to scholarships that might assist them in the admissions process at SAIC.

GTreasury has a powerful platform at the core of its business, enabling its global enterprise clients full visibility into their cash, payments and financial risk management. However, the company wanted to make sure their customers were receiving the best experience. They engaged SPR to help develop their vision for a new, high-quality platform design.

Railyards are busy hubs of activity where a constant flow of consumer goods, including new-car shipments, come in and head out to their final destinations. The sheer volume of traffic can be operationally challenging, particularly during peak times, when transportation providers are fighting to keep up with little bandwidth to optimize. REMPREX, a transportation logistics company, is working with SPR to leverage the flood of information collected by railyard tracking systems to create exciting, industry-forward platforms that equip transportation decisionmakers with data-driven insights.​

As consumer goods containers are transferred from trucks to train cars and vice versa, these data platforms surface information and insights about what’s going on in the yard in near real-time – to ensure maximum safety and efficiency. ​

Our client, an IoT company, focuses on smart building IoT through their IoT ingestion platform. Prior to the 2020 COVID-19 pandemic, the company began providing indoor air quality (IAQ) sensors as well. As businesses of all sizes became more concerned about IAQ, the company saw greater demand for their products, and a change in their customer base. The client’s new customers were smaller, with fewer employees and less time to manipulate a complicated dashboard solution. ​

To support our client through these changes, SPR:​

• Developed strategic recommendations for a simplified interface to appeal to the new customers. ​
• Created an intuitive plug-and-play dashboard​
• Build a multi-cloud solution to support customer data sources​
• Collaborated closely with the client to share knowledge, determine the direction of the product, and chart a new direction that’s agile, customer-driven and responsive. ​

When local research university UIC’s developer partner was unable to complete an app needed for a research study due to COVID-19 challenges, SPR quickly mobilized to finish development and prepare the app for launch. The UIC team was researching the use of mindfulness meditation to encourage emotional regulation and curb risky behaviors in arrested youth on probation in Chicago’s juvenile justice system.​

To help the research team create a successful app, SPR:​

• Met with the UIC team multiple times per week to keep the project on track​
• Implemented features like in-app analytics to provide the research team with the information needed for the study​
• Built in functionality to address challenges faced by the study’s population, including irregular access to smartphones, the potential of limited ability to charge devices, and the high incidence of using different devices to access the app. ​

With offerings for fire protection, life safety, and security as well as building automation and HVAC, Siemens Building Technologies succeeds or fails based on customer service. But field technicians were experiencing information overload. The quantity of email the Operations and Sales teams received made it difficult to identify essential information in a timely manner. To ensure their customers receive quality service, employees needed immediate access to information in their specific areas of responsibility. The company engaged SPR to develop and launch a mobile app for Android and iOS that could help the teams find information faster.​

As a result of developing the mobile app, SPR:​

• Developed a communications solution named Siemens Wire that enabled employees to curate information most relevant to them and reduce decision making​

• Integrated a custom robust publishing and management backend, allowing the company to streamline publishing and maintain content in a centralized cloud location​

• Created a beautiful, easy-to-use, and engaging user experience​

Aon, the world’s largest insurance brokering firm, has employees around the world. Aon employees needed a quick and convenient way to locate colleagues with precise expertise in a company of 50,000 people. SPR created a streamlined and intuitive PeopleFinder mobile app that offers secure access to employee data stored onsite. Employees can search for, read about, and contact colleagues around the globe who can help their clients be more successful​.

To enable greater scalability and flexibility,  SPR: 

• Extracted Aon’s data away from the business logic and putting it in a different tier, using smaller microservices as a layer between the data stored onsite and the receiving device ​

• Tested for compatibility using Xamerin Test Cloud to show that the app would deploy successfully on the majority of phones employees have. Xamerin made it possible to run automated compatibility tests whenever there were major changes throughout the project.​

The client is a membership organization that provides education, tools and publications to their 18,000 practitioner members. As a membership-driven organization, providing relevant services to its members is key to member retention. The client engaged SPR to develop a machine learning model to determine which members were likely to leave the organization or “churn.” By predicting members who were likely to churn, the client could reach out to key members and offer incentives for them to stay. 

The client was using an automated ML, but were looking for a trusted advisor to review the results, suggest additional techniques to improve their model, and perform an independent ML effort to validate the results. Churn models are a common application for ML, but our client faced a class imbalance issue in their data. Due to the COVID-19 pandemic, they provided free membership extensions in 2020—meaning there weren’t many examples of member churn to use to train the model.

We helped the client team understand, curate, prepare and cleanse their data. To create a model that would accurately predict churn and address the class imbalance issue, we weighted the data of members who churned more heavily. 

Outcomes:
SPR’s model estimated the likeliness of a member to churn with an accuracy of over 85%. Throughout the process, we were able to determine which factors led to member churn and which did not. Ultimately, the client was able to put the model into production and use it to identify at-risk members to re-engage. 

Siemens Building Technologies provides automation technologies and services. These building automation systems help to integrate and manage a number of different common building features from one centralized hub. in 2017, Siemens launched MindSphere, a cloud-based, open IoT operating system that connects products, plants, systems, and machines; it allows industrial customers to make decisions based on valuable data generated by the Industrial Internet of Things (IIoT). ​SPR worked with Siemens to implement an event-driven serverless architecture based on Lambda serverless compute and DynamoDB serverless storage. The platform's operational observability is built on CloudWatch and X-Ray. With the implementation complete, Siemens' product teams can now focus on delivering superior Smart Building hardware products and software user experiences opposed to dealing with low-level technical minutia day to day.

Ball Seed needed to optimize inventory and reduce waste by better predicting demand. Building upon earlier engagements with SPR, where Ball’s internal and external data sources were streamlined, the Big Data project pulled from new sales reporting information for predictive analytics