Fourth Chief Technology Officer Christian Berthelsen has more than 21 years of software engineering and management experience. He is also executive director of Fourth-Bulgaria. Fourth-Bulgaria is a partnership with New Bulgarian University that focuses on academic research, practical application of technologies, and innovation. He sat down for a Q&A to discuss machine learning in the hospitality industry, Fourth’s partnership with the university, and its investments in machine learning technology.
Q: For those who are unfamiliar, what is machine learning?
A: Machine learning is an application of artificial intelligence (AI) that provides systems the ability to learn automatically and improve from experience without being explicitly programmed. It focuses on the development of computer programs that can access data and use it learn for themselves.
On a basic level, machine learning is just a technique for predicting future outcomes based on past experience. It involves creating a model that “learns” from past data so it can make predictions about likely future events.
It requires good data, sophisticated mathematical models and a lot of computing power. Since we look after so much critical data for our customers, we are uniquely placed to develop machine learning models that can make their businesses more predictable.
Q: Fourth funded a Ph.D. program in machine learning at New Bulgarian University. What is the program working on and what do you hope to achieve?
A: We are building a long-term, sustainable partnership with a great university to further the research and investment in AI and machine learning. We are working together with the professors and students hoping to use research to develop flexible prediction models dedicated to the hospitality industry. Very little research exists in this field. There are huge opportunities to invest in and expand the application of AI. The program has only just started but we hope to get some academic input into our models in the coming year.
Q: Has Fourth made any other investments in the area machine learning?
A: This is part of Fourth’s multi-million-dollar, multi-year R&D investment program for Data Science in Hospitality. This will not only will look at machine learning, but also at combining many more data sources together for a single version of the truth.
We have formed a dedicated internal team for machine learning. They will be dedicated to refining our models and researching new opportunities for predictability across all our platforms.
Q: How can machine learning improve restaurant operations?
A: In general, the value is not in the forecasts themselves but in the downstream processes that consume them. For example, we could predict the likelihood people might be off sick or identify areas that may require extra attention during an inventory check.
Q: Is there a way to quantify or estimate the improvements machine learning can bring to the restaurant sector?
A: One of the main reasons Fourth chose to invest heavily in this area is because so little is known about what can be achieved, how much better or easier we can make the life of a manager or employee in this industry.
There’s a lot of anecdotal evidence and we have started to collect benchmarking data from a vast number of hospitality businesses to support the analysis and articulation of potential benefits. We have evidence that the current entry-level algorithms and machine learning tools we introduced into the staff scheduling processes have yielded significant benefits. For example, one UK chain of more than 600 outlets saved more than $5.5 million in one year alone by managing its labor with our labor productivity solution.
We know that the average manager spends eight to ten hours a week doing manual admin tasks in the back-office. Most of which could be reduced or entirely removed.
Q: How does Fourth use machine learning in its offerings?
A: We have developed a demand forecasting algorithm that uses past data to predict future sales. This takes into account a wealth of factors. For example, the weather or local events. This provides an accurate prediction of the unique demand patterns for every menu item in each restaurant or hotel.
This is an important prediction that has a number of applications for hospitality businesses. For example, our labor productivity module uses demand forecasting to help managers plan staffing level. Our inventory module can use it to help managers to manage their stock levels and prevent wastage or shortages.
Q: How soon can we expect daily restaurant operations to use machine learning?
A: Our customers already do! Many of our customers’ restaurant managers use our demand forecasting algorithms to plan their daily staff schedules.
Q: How will restaurants integrate it? What will it look like for restaurant operators/employers?
A: This is all about providing actionable insights to our customers. The challenge is to identify these insights and present them in a way that our customers can easily digest and act on. The details of the math and models will be hidden from them.
For example, we are currently building a suite of mobile tools for managers in restaurants and hotels that will help them with day-to-day tasks such as shift management, ordering and stock counting. The forecasts generated by our machine learning algorithms can be used to identify recommendations that are delivered directly to a manager. It will help them to be more efficient and drive improved margins across their business.
Florida voters shocked the nation last November when they overwhelmingly voted to gradually increase the minimum wage to $15 per hour over the better part of the next decade.
Amid the tumult of a chaotic global hotel market, brands of all sizes are reckoning with myriad unexpected factors — worldwide supply chain disruptions resulting in temporary inflation, nationwide labor shortages confining operations, and regional lingering restrictions limiting growth — each making the global recovery slower than expected.
Between payroll, inventory, reporting, and more, it’s easy for a manager to get trapped in the back office, sinking under paperwork. Here’s how to throw your best steward a life jacket.