AI in Sweden

Trustly, the Swedish pay-by-bank payment processor serving 110 million consumers in 30+ countries, migrated from a legacy BI tool to Amazon Redshift + Amazon QuickSight + Amazon Q in QuickSight. The new analytics stack scaled users from 80 to 300, generates insights 4x faster, and costs 50% less to maintain. Over 6 months users across risk, data science, executive, product, and sales teams submitted 15,000 natural-language questions via Amazon Q with 89% accuracy.

Vattenfall deployed AI-equipped drones at DanTysk and Sandbank offshore wind farms and the 12-meter Blue Essence uncrewed vessel at Hollandse Kust Zuid, with Fugro reporting the uncrewed vessel emits 95% less CO2 than traditional survey vessels.

How AI helps Sollentunahem deliver faster, more accessible service Book a demo Platform Overview Integrations Chat agent Email agent New AI insights EbbotGPT™ Read article Building trustworthy AI agents in the EU AI Act era Solutions Telecom E-commerce Financial services Energy Public sector Travel

Klarna's OpenAI-powered customer service assistant, launched in February 2024, automated 67% of support chats and absorbed the workload of roughly 700 agents, saving an estimated $40M at launch and around $60M by Q3 2025, before a 2025 walk-back that reintroduced human agents for disputes and complex cases while keeping the AI handling 853 agents' worth of tier-1 work.

Hexagon partnered with AWS to automate Standard Operating Procedure document digitization using generative AI, achieving above 90% accuracy while significantly reducing manual effort and cost compared to baseline.

Klarna deployed an AI-powered customer service assistant that handles 66% of all customer service chats, equivalent to 700 full-time employees. The AI agent autonomously resolves queries related to order tracking, payment plans, refunds, and account management across Klarna's 150 million active consumers and 500,000 merchant partners in 45 countries.

Spotify, headquartered in Stockholm, operates one of the world's largest AI-powered music recommendation systems serving over 600 million users globally. The recommendation engine combines collaborative filtering, natural language processing, and audio feature analysis through deep neural networks to personalize playlists such as Discover Weekly and Daily Mixes for each user. Spotify's ML infrastructure processes billions of user interactions and listening histories to generate real-time recommendations updated every 24 hours. Audio analysis models use convolutional neural networks to extract features directly from raw audio waveforms, enabling the system to recommend new songs with similar acoustic characteristics even before they accumulate listening history.