Urgent.News

One page, thousands of outlets. See who else covered it.

Editions

AI

Start-up-Check: Wie diese Start-ups KI im Mittelstand erst praxistauglich machen

Weil KI-Modelle überzeugend klingen und trotzdem falschliegen können, setzen Anbieter wie Amber auf Kontext. Das überzeugt Kunden wie Ritter Sport und bringt dem Start-up erneut Millionen.

Start-up-Check: Wie diese Start-ups KI im Mittelstand erst praxistauglich machen

Artificial intelligence (AI) may sound impressive, but can still lead to incorrect business decisions. This discourages companies, especially small and medium-sized enterprises, from adopting generative AI due to the fear of accepting false information and a lack of technical expertise to ensure reliable operation. Start-ups such as Amber Search and Paretos address this issue by introducing an additional context layer between company data and language models.

This layer is meant to clarify industry-specific terms, figures, and decisions, resulting in more reliable outcomes. Amber has attracted customers like Creditreform, Ritter Sport, and Zentis and has raised another seven million euros from investors. According to Bastian Maiworm, co-founder and Chief Revenue Officer of Amber Search, "58 percent of Fortune-500 companies use GenAI company-wide, while only four percent of the German medium-sized enterprises do."

Maiworm aims to bridge this gap and build trust in the technology among small and medium-sized enterprises in the German market. Amber was founded in 2021 by Bastian Maiworm, Philipp Reißel, and Igli Manaj in Aachen, where the three met during their studies at RWTH Aachen.

Written by urgent.news from Handelsblatt's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at handelsblatt.com →

More in AI

Zhipu va ouvrir les poids du meilleur chasseur de failles

L'essentiel Zhipu AI a publié GLM-5.3, entraîné sur un modèle de base de 743 milliards de paramètres et présenté comme dédié à la cyberdéfense et à la programmation.

  • Zhipu AI unveils GLM-5.3, a 743 billion parameter model for cybersecurity and programming.
  • GLM-5.3 scores 84.5 on CyberGym, outperforming Mythos 5 and GPT-5.6 Sol.
  • Zhipu will release model weights in stages after security evaluations.

The 7 AI Repositories I Starred This Month

I don't star GitHub repositories just because they are popular. A repository earns a star from me when I can see myself returning to it later. Maybe it solves a real engineering problem.

More from Monday 17 August →