Swedish BNPL (buy-now-pay-later) giant Klarna made waves when it decided to transform its Customer Success function - moving from human to hybrid. Here’s a quick recap of what they did👇

<aside> 💡

Klarna is a fintech that offers "buy now, pay later" (BNPL) services, allowing consumers to make purchases without upfront payment. It acts as a lender to consumers, and provides payment gateways, processing services and APIs to merchants.

</aside>

Background

Prior to its AI implementation, Klarna's workforce was huge - it employed around 5,000 people in August 2023. Its Customer Success **team alone consisted of 3,000 employees (~60% of its entire workforce), spread across India and the Philippines. The high costs of human capital paired with the costs of hardware and facilities management, meant that Klarna was operating on tight margins.

The AI Replacement

Spotting the reasoning and natural language processing (NLP) capabilities of its ChatGPT models, Klarna built an OpenAI-powered customer success chatbot [trained on Klarna’s internal wikis, how-tos and guides] to manage a wide range of customer queries, including:

• Customer queries
• Refunds and returns
• Invoice inaccuracies
• Real-time updates on balances and payment schedules
• Advice on spending limits and purchasing options