
Klarna, the global payments and shopping service, is undergoing a radical transformation, aiming to become what its CEO Sebastian Siemiatkowski envisions as the most efficient AI-driven bank. This isn't just about bolting AI features onto existing products; it's a fundamental restructuring of how the company operates, with a core principle that mirrors "eating your own dogfood." Klarna is aggressively developing and deploying its own AI-powered solutions internally, with widespread adoption across its workforce. This deep, internal immersion in AI is not only reshaping Klarna's operations but also providing invaluable insights that influence its technology and product strategy.

### An AI-First Culture: "Kiki" and Beyond

Klarna's commitment to internal AI adoption is striking. As reported by [CX Network](https://www.cxnetwork.com/artificial-intelligence/news/klarna-closer-to-ipo-as-ai-and-automation-boost-pays-off) and other outlets, a significant majority—reportedly over 87% to 90%—of Klarna employees utilize generative AI in their daily work. This isn't limited to technical teams; departments like communications (93% adoption), marketing (88%), and even legal (86%) are leveraging these tools.

A key enabler of this widespread internal adoption is "Kiki," Klarna's bespoke internal AI assistant. Kiki handles around 2,000 employee inquiries a day, covering everything from project statuses to procedural guidelines, effectively becoming a central nervous system for internal knowledge and support. This extensive use of an internally developed AI tool showcases a powerful "dogfooding" approach. By building and relying on Kiki, Klarna's teams directly experience the benefits and limitations of AI assistance, providing a rich feedback loop for its refinement and for understanding how AI can best augment human workflows.

### Replacing External SaaS with Internal AI Solutions

Perhaps one of the boldest moves in Klarna's AI strategy is its decision to phase out numerous external Software-as-a-Service (SaaS) platforms, including major enterprise tools for CRM, HR, and payroll. As [The Asian Banker](https://www.theasianbanker.com/updates-and-articles/klarna-s-ai-revolution-the-future-of-an-ai-driven-bank) notes, Klarna is "consolidating operations into a unified AI-powered ecosystem" by replacing off-the-shelf software with in-house AI systems optimized for automation, data standardization, and efficiency.

This strategy is rooted in the Pareto Principle—the idea that 80% of users typically rely on just 20% of a software's functionality. Instead of paying for "bloated SaaS solutions," Klarna is developing lean, purpose-built AI systems that cater to its core needs. This massive internal development and deployment effort means Klarna is its own first and most demanding customer for these critical operational platforms. The engineers building these systems receive immediate feedback from their colleagues across the company who depend on them daily. This internal pressure cooker drives rapid iteration, efficiency, and a focus on genuine utility.

### AI in Customer Service: A Real-World Proving Ground (with Lessons Learned)

Klarna made headlines with its AI-powered customer service assistant, which, in its initial rollout, reportedly handled two-thirds of all customer service chats—the equivalent workload of 700 human agents—and significantly reduced resolution times. This was a large-scale deployment of an AI system directly interacting with Klarna's core user base, with Klarna's own operational metrics serving as the primary measure of success and areas for improvement.

Interestingly, this aggressive AI-first approach in customer service has seen some recalibration. As reported by [Cresta](https://cresta.com/news/klarna-shifts-focus-back-from-ai-to-human-interaction/) and [FinTech Weekly](https://www.fintechweekly.com/magazine/articles/klarna-hires-customer-service-after-ai-pivot), Klarna is now resuming hiring for human customer service agents. CEO Sebastian Siemiatkowski acknowledged that an overemphasis on cost-cutting through AI may have impacted service quality. This adjustment is, in itself, a testament to the "dogfooding" process—Klarna experienced the real-world outcomes of its AI deployment at scale, identified limitations (likely through both customer feedback and internal monitoring using its own analytics), and is now iterating to find a better balance between AI efficiency and human-centric support. This learning phase is critical and showcases a responsive approach to internal and external feedback.

### Shaping Consumer-Facing Products

While the most prominent examples of Klarna's internal usage focus on operational AI, the insights gained inevitably flow into their consumer-facing products:

* **AI-Powered Shopping Assistant:** Klarna has integrated AI into its consumer app to offer features like personalized product recommendations, product comparisons, and price searches across merchants. The extensive internal use of generative AI by Klarna employees, including understanding its capabilities for search and summarization through tools like Kiki, would directly inform the development and refinement of such consumer AI features.
* **Risk Management and Fraud Detection:** Klarna's core Buy Now, Pay Later (BNPL) service relies on sophisticated AI and machine learning models for real-time credit risk assessment and fraud detection. These models are continuously trained and refined using vast amounts of transaction data. While not "dogfooding" in the sense of an employee using a UI, the internal data science and risk teams are constantly interacting with, evaluating, and improving these critical AI systems that are foundational to Klarna's business.
* **The Klarna App and Card:** With a large employee base, many of whom are likely users of Klarna for their own shopping and payments, there's an inherent, ongoing internal user study. Feedback on the app's usability, new payment features, budgeting tools, or the experience of using the Klarna Card can provide valuable qualitative data to product teams. While Klarna provides [testing environments and sample data for merchants](https://docs.klarna.com/resources/developer-tools/testing-payments/before-you-test/), its own employees navigating the live consumer ecosystem offer a different, yet equally important, layer of real-world feedback.

### Benefits and the Path Forward

Klarna's deep internal reliance on its own (often AI-driven) technology provides several key benefits:

* **Rapid Innovation in AI:** By being its own primary developer and user of AI operational tools, Klarna can iterate and innovate in this space much faster.
* **Deep Understanding of AI Implementation:** Klarna is gaining unparalleled experience in the practicalities, challenges, and benefits of deploying AI at scale across an entire organization.
* **Efficiency and Cost Savings:** Successful internal AI tools have already demonstrated significant cost reductions and productivity gains.
* **Potential for Unique Product Offerings:** The knowledge gained from building and running an AI-driven internal ecosystem can translate into unique and powerful features for both consumers and merchants.

However, this approach also comes with the responsibility of maintaining and securing a growing suite of proprietary software, as noted by The Asian Banker. The recent recalibration in AI customer service also shows that the journey involves continuous learning and adaptation.

Klarna's bold strategy of building and extensively using its own AI-powered tools for internal operations positions it as a fascinating case study in modern "dogfooding." It's a company that is not just adopting the AI revolution but is fundamentally rewiring its internal workings with it, using its own employees and operations as the primary proving ground for the future of its financial technology.

Articles tagged with "Klarna"

Klarna's AI Revolution Starts Within: How Internal Adoption Fuels a Fintech Powerhouse