Klarna CEO Replaces SaaS with AI

Klarna's AI-powered customer service assistant is doing the work of 700 full-time agents and is projected to drive a $40 million profit improvement. In its first month, the assistant handled 2.3 million conversations, which accounted for two-thirds of all customer service chats. This has reduced errand resolution time from 11 minutes to under 2 minutes and led to a 25% decrease in repeat inquiries. The company has also seen a significant impact on its marketing operations, cutting sales and marketing spending by 11% in the first quarter of 2024. This resulted in an estimated $10 million in annualized savings, with $6 million from reduced image production costs and $4 million from lower spending on external marketing agencies. Klarna now uses generative AI tools like Midjourney, DALL-E, and Firefly to create images, reducing the production timeline from six weeks to seven days. Internally, Klarna has developed over 300 GPTs for employee use, and an internal AI assistant named "Kiki" has answered over 250,000 employee inquiries. The company's adoption of AI extends to copywriting, with an internal tool called "Copy Assistant" handling 80% of this work. While initial reports suggested a complete replacement of SaaS platforms with a single AI, CEO Sebastian Siemiatkowski clarified the move was more about consolidating fragmented data from over 1,200 SaaS tools into a more unified internal tech stack. This was to create a centralized "hub of knowledge" to improve the effectiveness of their AI models. The company did replace some SaaS vendors with others, like moving to Deel for HR functions. This strategic shift towards in-house AI development is part of a broader goal to become a more efficient, AI-driven bank. The company's workforce has been reduced from a high of around 7,000 to approximately 3,000, and the CEO anticipates the number could fall below 2,000 by 2030 through natural attrition and reduced hiring.