Aurora, as the AI assistant is called, is based on citizens and their real-life situations and needs. Aurora is meant to facilitate smooth and convenient service consisting of solutions from various providers – exactly the right service at the right time. The client, the Finnish Ministry of Finance, and the project are aiming high. Instead of settling for an AI device (that would rapidly become obsolete), they developed an entire network of intelligent services. These are services that public and private providers, and perhaps even individuals, can create together.
Advanced machine learning
Aurora is based on machine learning (specifically, reinforcement learning), where the software combines services from public and private providers that get the highest reviews by users over time. The user groups are created based on anonymised personal data.
“The more anonymous information we have at the group level, the greater the potential to offer relevant services at the individual level. As a result, we can bring the citizen and society closer together,” explains Antti Hahto at Knowit Finland.
Tested on real-life events
The feasibility study was carried out based on three typical real-life events: a move to a new town to attend university, skills development to improve employability and support to children and parents when family relationships change.
“By using information and AI, we can create new types of services that cross administrative lines. Focus is on the citizen’s needs and perceptions of smooth and convenient service, not the needs of an individual government agency,”
says Minister of Local Government and Public Reforms Anu Vehviläinen, who is responsible for the digitalisation of public services.
The feasibility study will be completed and reported in spring 2019. If it proves successful, further development will follow. Something to look forward to, without a doubt!