For several years now, technology has been changing due to a deep digital transformation. As a result, FuturMaster has moved towards a more cloud native technology with a micro-service architecture that allows us to develop new features and get quick feedback from customers.
Agile methodology: Users at the heart of our development process
The agile development methodology was adopted by FuturMaster to ensure that our innovations are in line with the needs of our users and customers. We collaborate with them especially on the design of the functionalities and they have the possibility to test them, which also enables us to guarantee constant improvement of our solutions. With this agile methodology, we can also focus our efforts on delivering value to our customers by delivering increments and collecting feedback from our users once again to facilitate the prioritisation of topics. Thanks to this method, we are now able to continuously deliver new products.
Business process management
Business Process Management (BPM) will allow us to describe all the processes within a company and how these processes will interact with each other. We are currently working on integrating these tools into FuturMaster to manage:
- The management of a business workflow: this tool will be useful, especially when one is projected on a promotion sheet, by accompanying and managing the different stages in the validation of promotions. This type of tool will describe the validation process and allow each user to validate the promotion.
- The triggering of processes interactively: It will allow us to trigger different treatments interactively, especially on promotions. Thanks to the Business Process Management tool, we will be able to trigger processing (creation of nodes, quantity splitting, etc.) when the promotion will be retained by the retailer.
- Periodic treatment orchestration: This tool will apply to night, weekly and monthly treatments. With this tool available in FuturMaster, we can orchestrate different treatments and have direct access to all the calculation services from the configuration of these processes. In order to ensure that these daily processes go well, the tool will be fully integrated with error management and monitoring cockpits.
Key development initiatives: Intelligence/ Machine learning
We deploy several methods using Machine Learning: regressive methods and methods using neural network that will be adapted over the short, medium or long term.
- Automated cleaning: Thanks to a machine learning algorithm, we will be able to carry out automated cleaning of the history. According to defined sensitivity parameters, we will specify potential outliers that are likely to correspond to exceptional events that will have to be cleaned. Automated cleaning will optimise the time of the demand planner, especially at the beginning of the project.
- Demand sensing: We have at our disposal more and more accessible data that will potentially influence demand in the future. It can be both traditional data, competitive market data, but also social data. The forecast calculation is based on the latest sales information as well as external variables. The main objective is to take advantage of the data set to continuously improve the sales forecast.
- New Product Launch: We are implementing a new calculation approach that allows a better sales forecast on product launches by anticipating sales volumes and further customising. In addition, a new Machine Learning approach will consider all the launches made with the corresponding sales volumes as well as the characteristics of these launches and will make it possible to classify these by large types of sales profile to predict future launches.
- Supply Planning: Due to the proliferation of decision-making issues related to the Supply Chain network and the complexity of the network involving several intermediaries, efficient tools are needed. Consumers expect an ever-higher rate of service. In order to respond to this, we are working on the notion of resilient planning that allows us to make simulations considering the uncertainties of demand, distribution, supply and production.
- New user experience: We are also implementing the next generation of FuturMaster products that rely on the development of new user interfaces to support the digital transformation of our customers by equipping them with analytical capabilities and more automatic and dynamic decision-making. To succeed, we deploy easy-to-use solutions for users and we also work on data intelligibility to help decision-making.
In 2019, we developed the use of machine learning for new product launches and all the models for demand sensing.
Additional modules were made available at the end of 2019 and will be released in 2020. This includes the new Dashboard, TPx (Trade promotion management), the integration of the BPM tool to manage workflows. In fact, this module consists of a new user interface including predictive algorithms with Machine Learning approaches as well as the valuation and calculation of pre-packaged ROIs by default.
The new Dashboard will allow users and customers to be at the heart of navigation. In addition, work is also under way on a new demand management module with a completely revised user interface, from the modelling sheet to the new Machine Learning algorithms. This new simplified and modern web interface guarantees more possibilities in terms of calculations.
Working on Machine Learning approaches also allows us to recalculate a plan quickly to respond to changes. In the same way, we will focus on the next generation of the Supply Planning module, developing a new user interface for parameter management.