Session Outline
Skatteetaten, the Norwegian Tax Authority (NTA), has one of the highest digital maturity levels amongst the public organisations in Europe. The agency collects and administers some of the largest datasets available in the country, collected from individuals, firms, public institutions, agencies, and registries. In 2022 the agency explicitly incorporated the use of machine learning and AI in its strategic ambitions, creating a big transformational journey. In this presentation, we will give an overview of how the Norwegian Tax Authority is putting this strategy into practice. We present our approach for building and implementing large scale ML models in a cross-disciplinary, complex, and fragmented setting, bringing examples from the case of the NTA’s risk scoring models for companies. The presentation will shed light on common organisational, technical, and capability challenges for developing AI-enabled models and processes in Nordic public organisations, and provide resolutions to unlock value at scale.
Key Takeaways
- Put together a cross-sectional team, with represented all needed expertise in the entire value chain.
- Data preparation is the absolute priority. The value is 80% in the data, and 20% in the models.
- Create an environment where as much as possible can be reused or is useful transversally in the organisation.
