Successful Presentation of Aisio.Ai on an International Stage

On Tuesday, April 29, 2025, we had the opportunity to present our document analysis platform aisio.ai to an international audience, together with our technology partner deepset.

In an engaging presentation, Ralf Kauther (credX) demonstrated the possibilities that arise from using generative AI for analyzing unstructured documents. Using practical examples, he showed how aisio.ai significantly simplifies information extraction from complex texts – thereby considerably accelerating decision-making processes in commercial real estate financing.

Following this, Laura Edinger-Schons from deepset provided insights into the underlying technical architecture. She explained how targeted system optimization allowed for excellent results in the quality dimensions crucial for us: correctness, source attribution, and completeness. The outlook on future developments was particularly exciting: With the help of an agentic workflow, we aim to further automate and improve data analysis.

Our sincere thanks to Laura, the team at deepset, and all attendees for their great interest and valuable feedback!

Insights from the Webinar

During our webinar, we conducted a non-representative flash survey among participants to gauge sentiment regarding the practical application of LLMs (Large Language Models):

  • 60% of participants are currently still in the exploration phase and are evaluating potential use cases.
  • 20% are already planning a concrete implementation.
  • 20% already have a solution in production.

When asked about the biggest challenges in implementing productive AI applications, participants responded:

  • 50%: Technological Challenges (e.g., infrastructure, model performance)
  • 50%: Identification of Relevant Use Cases and Collaboration Between Humans and Technology

These results show: The use of generative AI is still in its early stages in many organizations – the potential is vast, but practical implementation requires a strategic approach and suitable partners.