t3n Arbeit in Progress

Use Case: Wie nutzt ein Kosmetik-Unternehmen KI?

36 min

This episode of t3n Arbeit in Progress explores how the cosmetics company Kosnova integrates generative AI into its operations. Guest Constanze Stein, an Expert Gen AI, discusses the company's decentralized approach to AI adoption, emphasizing user-centric training and specific use cases like the 'Product Concept Bot' to streamline product development. Key insights include the importance of setting secure frameworks, leveraging enterprise tools like ChatGPT, and engaging leadership to drive transformation without creating separate AI initiatives.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Kosnova employs a decentralized AI adoption strategy, empowering teams to experiment with tools like ChatGPT Enterprise within a secure framework, rather than imposing top-down mandates.
2

The 'Product Concept Bot' was developed through close collaboration with product marketing teams, addressing their need for structured concept development over pure creativity, highlighting the value of user-centric design.
3

The company focuses on integrating AI into existing leadership and HR trainings, such as feedback culture, rather than creating standalone AI courses, to embed AI into daily workflows.
4

Constanze Stein emphasizes that AI should not be scaled through prompts or prototypes alone, but through collaboration and cross-team efforts to truly transform work processes.
5

Kosnova plans to move from numerous similar AI assistants to a more integrated system of agents connecting different departments, aiming to achieve this within one to two years.

💬 Notable Quotes

"We work to rationalize ourselves away. I don't want a team that's only busy using Excel; I want AI to arrive as a tool, as a quality of digital transformation in the company."
"We don't scale AI through prompts or the next cool prototype, but through collaboration. We won't get to the big cases or real transformation without working closely together."
"The biggest challenge is almost the expectation management. You get all these market impulses about companies generating millions with agents, but then you have to consider what it means to implement that in an audit-safe, legally compliant way."