Press & press releases

German startup deAI Labs GmbH is launching the AI platform daeo. daeo connects scattered enterprise data into controlled enterprise context and makes it usable across different AI models, applications and interfaces. The platform is already being used successfully in pilot projects in manufacturing, the service sector and the public sector.

Tutzing, 6 October 2026 – deAI Labs GmbH, based in Tutzing near Munich, is today launching the sovereign AI platform daeo. daeo connects scattered data from ERP systems, databases, documents and tickets into controlled enterprise context and makes it available to a wide range of AI models, applications and interfaces. The platform was built for medium-sized businesses and the public sector. Demand is high: according to the ifo Institute (June 2026), only 47.2 per cent of medium-sized companies in Germany use AI – across all company sizes, the figure is 54.5 per cent.

A key bottleneck is the underlying data: according to Gartner, 63 per cent of organisations lack suitable data management practices for AI or are unsure whether they have them. According to a recent study by Lünendonk, 95 per cent of companies cite insufficient data quality or availability as a barrier to the success of their agentic AI projects.

daeo was developed under the leadership of deAI Labs co-founder Lukas Beckenbauer, who previously conducted AI research at the Technical University of Munich and the University of Cambridge. The name daeo stands for Data Extraction for Organizations. The platform is being used successfully in extensive pilot projects across three sectors.

Why AI projects fail because of their underlying data

There are AI models for a wide variety of use cases. But their use is held back by the data – more precisely, by its preparation. In February 2025, Gartner predicted that by the end of 2026, around 60 per cent of AI projects without AI-ready data would be abandoned. At the same time, companies' IT landscapes are fragmented: according to a Forrester survey (August 2026), only seven per cent of ERP decision-makers operate a single ERP instance. Enterprise AI therefore needs to work across system boundaries. daeo helps by preparing data from any source for AI in a context layer.

Context is more than access to data. An AI model that finds a document does not yet know whether it is the valid version, which process it belongs to or who is allowed to see it. These relationships are not contained in any dataset – they are embedded in established rules, responsibilities and employees' knowledge gained through experience. This problem cannot be solved by a different AI model, but by optimally preparing the data in a specialised context layer.

deAI Labs' premise is that data onboarding – the journey of heterogeneous enterprise data distributed across systems towards a state in which AI can reliably find, understand and use it – will become a decisive factor in 2027 in whether AI projects also work in medium-sized companies.

How the daeo Context Engine works

The daeo Context Engine builds a context layer across a company's IT systems – ERP, SQL databases, MES, documents, tickets, email and collaboration systems. The platform makes scattered content accessible and connects data relationships, access rights and knowledge into controlled enterprise context with source references and role-based access. This context updates automatically in real time and is available through the daeo App, MCP and API for different applications, models and AI agents.

The architecture is modular and designed for rapid deployment: a first clearly defined use case can start with standard sources within 48 hours. More complex data and ERP landscapes are added step by step. Sources, permissions and relationships already made accessible are reused for further applications. The system actively updates itself and adapts more closely to company-specific requirements over time.

Why sovereign preparation of enterprise context matters for a flexible AI strategy

In an analysis from August 2026, Forrester identifies a key question: whether companies can carry not only their data but also their definitions, relationships and permissions across systems, models and interfaces. daeo enables companies to make all their data accessible for AI while remaining independent of individual system providers or AI models.

The founders of deAI Labs believe that EU hosting is only the beginning of AI sovereignty. In their view, sovereignty also means the freedom to switch between AI models at any time. With daeo, the AI model can therefore be freely changed depending on the use case. The enterprise context is retained. Customers also choose their own level of sovereignty in other respects: daeo can run in European data centres, in a private cloud or on their own infrastructure.

Successful pilot projects in manufacturing, services and the public sector

Before the market launch, daeo was rolled out in extensive pilot projects with customers in manufacturing, the public sector and the service sector. One already documented example from the public sector is software provider komuna GmbH, which provides software for around 700 municipalities: at komuna, daeo makes the wealth of experience from 111,000 historical tickets usable for support staff. Similar previous cases can now be found in seconds. According to the customer, daeo delivers “huge time savings” in the day-to-day work of the company, which answers an average of 200 support requests a day. It took only three months for the platform to enter productive use, and komuna intends to use it for other use cases in the future.

The founding team: AI research meets venture capital experience

Lukas Beckenbauer (CEO) is responsible for daeo's overall strategy, vision and corporate management. He conducted research into artificial intelligence and distributed systems at the Technical University of Munich and at the University of Cambridge's Supply Chain AI Lab. His research on self-optimising AI agent systems and decentralised organisational structures has been presented at leading international forums such as NeurIPS and the Academy of Management. He is also a lead author of the study “Synergies of Blockchain and AI – Vision of a European Data Economy”, published by the Konrad-Adenauer-Stiftung, in which he argues for European technological sovereignty and decentralised AI infrastructures. His commitment to combining sovereign, diverse AI with security and governance shapes daeo's architecture.

Sarah Gottwald (CGO) is responsible for daeo's operations, business development and strategic partnerships. She brings extensive venture capital experience – as Managing Director of Blockchain Founders Group and from her work at Vanagon Ventures – and has been a member of the supervisory board of NYALA Digital Asset AG since 2025. Before working in venture capital, she spent nine years at BMW in finance and strategy roles focusing on autonomous driving, where she negotiated partnerships and completed merger and acquisition transactions.

Availability

daeo is available from 6 October 2026. Adoption is modular, beginning with a first use case and then expanding to additional workflows, teams and systems. Further information, the product overview and the komuna GmbH case study are available at daeo.ai.

About daeo

daeo is a sovereign AI platform for medium-sized businesses and the public sector. The daeo Context Engine processes scattered data from ERP systems, databases and documents largely automatically into controlled enterprise context with source references and role-based access. The enterprise context is fully auditable and updates automatically. AI models, data sources and interfaces can be flexibly chosen at any time. daeo runs in European or German data centres, in a private cloud or on customers' own infrastructure; users decide their own level of sovereignty. daeo is a product of deAI Labs GmbH, based in Tutzing near Munich. Further information is available at daeo.ai.

About deAI Labs GmbH

deAI Labs GmbH is a research and innovation hub for artificial intelligence based in Tutzing near Munich. Founded in April 2026, the company combines current research into AI systems, federated data infrastructures and the coordination of distributed actors with the development of practical enterprise software. Founded by Lukas Beckenbauer and Sarah Gottwald, deAI Labs is developing daeo as the context infrastructure for European companies that want to combine productive AI with long-term technological freedom of choice. Further information is available at deai-labs.com.