News | Munich | 28.01.2025

AI that does what it's supposed to: moresophy applies for a patent for DAPHY®

Patented hybrid technology combines precision, efficiency and control for SMEs

Munich, January 28. January 2025 – Munich-based moresophy GmbH has been granted a US patent for its innovative hybrid AI technology. DAPHY® registered. Specially developed for medium-sized companies, DAPHY ® companies a secure, efficient and targeted use of generative AI, without high costs or complex learning curves. The technology lowers the barriers to using AI and helps to optimize business processes based on data.

Many SMEs recognize the potential of generative AI, but are reluctant to implement it. A lack of expertise, high costs and uncertainties when monitoring the results are among the most common inhibitions. This is where moresophy comes in. For over 20 years, the Munich-based company has been developing practical AI solutions that combine transparency and efficiency.

With DAPHY®, we want to break down the barriers for small and medium-sized companies when using AI and offer a solution that is easy to use, works precisely and creates tangible benefits in day-to-day work,” explains Prof. Dr. Heiko Beier, founder and CEO of moresophy. “DAPHY® is like a reliable partner: it understands the specific requirements of a company, delivers consistent results and increases efficiency.

As early as 2024, moresophy received a patent for a resource-efficient technology for processing unstructured data – the basis for expanding the portfolio with the hybrid AI DAPHY®.

(AI-based) assistance for prompting the AI

DAPHY® combines the creative power of generative AI with the analytical precision of machine learning algorithms. At its core, the technology intelligently prepares company data and uses it to control generative AI models in a targeted manner. This enables companies to perform a wide range of tasks efficiently and reliably: from automatically adapting product descriptions to different target groups and generating precise, context-related answers in customer service to supporting well-founded business decisions by combining qualitative assessments with quantitative key figures in reports, studies or contracts.

A decisive advantage of DAPHY® lies in the easy handling and adaptability of AI to specific requirements by end users. The quality of the results of generative AI in the form of LLMs depends largely on how specifically it understands the context of existing data. This can be achieved in various ways. Training new LLMs with comprehensive data is extremely resource-intensive and incurs costs in the multi-digit million range. Pre-trained LLMs can be optimized for specific tasks or specialist domains by means of fine-tuning. However, this is also reserved for specialists and also requires expensive GPU resources. For this reason, so-called retrieval augmented generation (RAG) approaches, which apply the context understanding trained in the LLM to the user’s own data, have proven their worth. With the disadvantage that the LLM works as it it this deems appropriate.

Intelligent, automated prompting with DAPHY®

DAPHY® on the other hand, relies on the intelligent generation of prompts using structured data. The prompts are precisely optimized to the relevant context in real time. This makes prompting automated, but at the same time transparent and controllable for the user. The often time-consuming process of prompt engineering is eliminated and the results are not only precise, but also traceable and reproducible at all times.

Since DAPHY® works with structured data obtained from a company’s raw data and information, the results can be verified by referring back to the sources. This increases reliability and minimizes the risks of AI-generated misinformation.

In addition, DAPHY® saves resources and can be easily integrated into existing IT structures. Companies can use their own data directly and remain independent of costly cloud solutions – without relinquishing control over sensitive information.

With DAPHY® we show that generative AI can not only be powerful, but also controllable and sustainable,”says Prof. Dr. Heiko Beier. “You don’t need a data center the size of a nuclear power plant to put AI to good use.”

Hybrid AI with maximum control at minimum cost: DAPHY® combines efficiency and precision where other AI approaches reach their limits.
Hybrid AI with maximum control at minimum cost: DAPHY® combines efficiency and precision where other AI approaches reach their limits.

Patent status and outlook 

The patent application for the DAPHY®-technology was filed on December 19, 2024 and is currently in the examination process. The company plans to use the patent to drive forward the further development of hybrid AI technology and expand its use in various industries. The solution should be even more flexible and easier to integrate into existing processes in future.

About moresophy 

moresophy is an innovative provider of cutting-edge AI solutions that helps companies to precisely analyze unstructured data and automate business-critical processes. With over 20 years of experience and patented, GDPR-compliant technology, moresophy offers customized AI solutions that can be seamlessly integrated into existing data environments and business processes. The platform combines context-sensitive data analytics, machine learning and AI methods to achieve reliable, explainable results and ensure the protection of data and intellectual property. Companies benefit from increased productivity to strengthen customer relationships and realize data-driven business models.  

Further information can be found at www.moresophy.com 

Hybrid-KI mit maximaler Kontrolle bei minimalen Kosten: DAPHY® vereint Effizienz und Präzision, wo andere KI-Ansätze an ihre Grenzen stoßen.
Scroll to Top
Cookie Consent Banner by Real Cookie Banner