Dynamic multivariate optimisation model based mobile-app using autoML machine learning algorithm made for shift workers in retail

Shift management is complex and is subject to many constraints and interdependencies. Shift working is utilised across industries as large as the NHS to the retail, military, government, and many other sectors. Current solutions only provide tools to place shifts on calendars; they do not address the complexities directly, instead relying on human intervention to provide the decision-making and management components e.g. moving a shift around.

We intend to solve for shift working patterns by applying a linear optimisation model, similar to how the travelling salesman problem was used to solve distance travelled by a sales rep and hence massively reduce carbon footprint and cost to the employer, our project will create software that combines a mathematical approach that is impartial, eliminates subconscious biases, uses machine learning algorithms (autoML) and live data to dynamically update the linear optimisation model, enabling AI-generated solutions to the shift scheduling problem.

The project will thus enable software to replace excel and other rudimentary gut feel rota schedules where thousands of people are involved, enable software processes to interact to request shift changes, update hours, make more or fewer hours available, switch shifts, and deploy the optimal staff combinations for the best business outcomes.

Replacing the person responsible for shift scheduling with an AI-generated solutions will be akin to removing the taxi dispatcher, the way that Uber has done with technology. Anticipated benefits from the project include greater efficiency and performance; reduced overheads; reduced stress and anxiety; improved service for customers and therefore improvements in reputation; and increased fairness (eliminating subconscious bias, for example).