6.4. The realisation of use cases on the data platform

Below are examples of how the previously examined three key use cases would be realised on the data platform. Before the building of an analytical model, the datasets related to the use case must be imported to the data platform, unless the data products needed for the use case already exist. Once the necessary datasets have been imported to the platform, cleaned and pre-processed as required, the data can be utilised in the analytics environment.

Improving the customer experience

In accordance with the City’s Digitalisation Programme, Helsinki is investing in proactive and individualised services. Proactivity requires the data platform to have the capability to predict a customer’s individual needs before the customer contacts the City.

From the perspective of the data platform, this means that customer-specific needs must be proactively anticipated based on data from an individual customer or customer group. When the algorithm developed for this purpose finds a service that matches the customer’s individual criteria, the service in question can be recommended during various customer encounters, such as in connection with a customer service transaction or by sending the customer a message about the service utilising the contact details available via their Helsinki profile and according to the communication preferences provided by the customer. From the perspective of the data platform, messages can be sent via batch processing or an API, which also makes it possible to send real-time messages directly from operative systems. Examples of this type of proactive service are the offering of a school place or language course to a customer based on their background information.

Assisting City personnel by optimising service processes

The services produced by the City generate large amounts of data, which can be collected and analysed to create data-based situation pictures and optimise service processes.

From the perspective of the data platform, this means that data is collected and combined from different information resources and published for visualisations.

The optimisation of service processes is carried out using optimisation and simulation algorithms, which identify the process options that best optimise the objectives set for the algorithm. The output of this optimisation can be, for example, an estimate of how many employees are needed at a service point at a specific time of day, which a team manager can utilise to prepare optimised work schedules.

Data-driven decision-making

The use case for data-drive decision-making involves combining data from several information resources to analyse the cross-effects of different phenomena. The development of analytics tools and methods can help diversify the range of information provided to decision-makers. In addition to utilising traditional statistics and historical data, decision-makers will be able to make use of datasets procured through new data collection methods and apply forecasting models to them. These models are created using machine learning methods, i.e. by teaching them with historical data. Once a forecasting model is taught, it can be used to produce forecasts and scenarios to support decision-making. The produced scenarios can also be visualised using dashboard indicators, for example. Combining datasets and information models with the city model of Helsinki will create a platform and digital twin that can be integrated into the City’s service processes to improve efficiency, productivity, operational planning, monitoring and transparency.