Four Domains.
One Thesis.
Industrial Intelligence
The Architecture of Institutional Capability
Where AI and institutional complexity intersect — and where the gap between what is technically possible and what organisations actually govern is widest. We have spent two decades working at that gap. What follows is where we work. Not a portfolio in the conventional sense. A thesis about where Europe’s technological future will be decided.
I
Software Productivity & System Governance
Software development is the largest unmanaged asset on the balance sheets of global corporations. Billions flow annually into code that no board can read, no CFO can measure, and no CIO can fully control. In any other capital domain, this would be considered a governance failure. In software, it has been normalised. Artificial intelligence now makes it possible to end that normalisation — to render the invisible legible, and to give leadership the instrument it has never had: real accountability over the systems it funds.
AI-driven analytics that make complex, distributed software ecosystems measurable and governable at the board level — in real time, across thousands of systems.
— Radical Transparency: AI aggregates millions of lines of code, repositories, and developer telemetry into hard financial and efficiency metrics. For the first time, the “black box” of software production becomes a governable asset class.
— Resource Governance: Pattern recognition at scale identifies toxic code, architectural debt, and systemic friction invisible to conventional review — stopping the waste before it compounds across the next development cycle.
— Board-Level Accountability: Governance structures established across hundreds of concurrent IT projects, without interfering with operational workflows — giving CIOs and supervisory boards the control layer their mandate has always required — and that regulators are beginning to demand.
II
Spatial Intelligence & Sovereign Digital Twins
The physical world now generates more structured data than any human team can interpret. Sensor arrays, satellite imagery, LiDAR networks, and infrastructure monitoring systems produce volumes that have long exceeded the analytical capacity of conventional methods. Machine learning resolves this — but only if the intelligence layer is owned, not rented. The question is no longer whether artificial intelligence reads the built environment. It already does. The question is who controls the system that does the reading.
Machine learning for automated classification and analysis of massive spatial datasets — building sovereign data infrastructures that no external platform can hold hostage.
— Sovereign Data Control: Spatial and geospatial AI infrastructures engineered for federal, state, and institutional deployment — structurally independent of foreign cloud platforms, foreign jurisdictions, and the political leverage they carry.
— Digital Twin Intelligence: Static asset portfolios — real estate, facilities, critical infrastructure — become continuously updated digital twins, processed by AI that sees what human inspection misses and responds faster than any audit cycle.
— Verified Reality: Executive decisions grounded in millimetre-precise, AI-verified spatial models rather than estimates and approximations — in domains where the cost of being wrong is measured in concrete, not spreadsheets.
III
Algorithmic Infrastructure & Digital Capital
The velocity at which capital markets now operate has structurally outpaced human decision cycles. Institutional actors that rely on legacy architectures are not simply slower — they are operating in a different temporal reality from the systems they interact with. Algorithmic and AI-driven infrastructure is no longer a competitive advantage. It is the baseline condition for operating at institutional grade. Those who do not control their own infrastructure do not control their own exposure.
Mission-critical algorithmic systems — engineered for institutional-grade operation under full regulatory scrutiny, at the speed and precision the next financial era demands.
— Institutional-Grade AI: Quantitative models and AI-driven analytics designed for high-frequency environments where error tolerances are set by regulation, not engineering preference — and where failure is a systemic event, not a product incident.
— Compliance by Architecture: Regulatory frameworks are not applied retroactively — they are embedded at the structural level of the system. AI and DLT components built to operate inside the regulatory perimeter from the first line of code.
— Infrastructure for the Next Decade: Critical financial systems designed today for the tokenised, decentralised markets of tomorrow — before the transition forces reactive adaptation under conditions that will not permit it.
IV
Industrial Edge & Visual Computing
Industry 4.0 made a promise it has not yet kept. The vast majority of industrial production environments still run on manual inspection, delayed reporting, and approximated quality control — not because the technology does not exist, but because the implementation path from research laboratory to factory floor has never been de-risked at scale. AI at the edge — processing image, video, and sensor data precisely where it is generated, without cloud dependency — closes the gap between what is technically possible and what operationally exists. The organisations that close it first will not merely be more efficient. They will be operating in a different industrial category.
High-performance AI for image, video, and multi-dimensional sensor data — processed at the point of generation, without latency, without cloud dependency, without compromise.
— Edge Intelligence: AI models deployed directly within production environments transform unstructured sensor streams into actionable operational insight in real time — eliminating the latency that makes cloud-dependent architectures unfit for industrial control.
— Scalable Deployment: Institutional-grade visual computing capability extended to production environments of all scales — removing the implementation barriers that have kept advanced AI confined to pilot programmes rather than operational reality.
— Autonomous Quality Assurance: Computer vision and multi-sensor AI replace manual inspection at the point of production — with a consistency, speed, and resolution no human process can match and no market pressure can erode.
The Foundation
Rigorous De-Risking
AI-based systems fail in critical infrastructure not because of capability gaps — but because of implementation risk. The transition from research prototype to industrial deployment is where most initiatives collapse. We have spent two decades understanding exactly why.
Every building block undergoes systematic maturation at the interface between our Institute and applied research — calibrated against industrial operating conditions, regulatory constraints, and organisational complexity, not laboratory benchmarks. The standard is not what is technically possible. It is what reliably holds under real load.
Maximum technological advantage. Minimum systemic risk.
Initiative
Software Excellence Network
A closed forum for senior technology executives and academic researchers working at the boundary of what European software infrastructure can do. The goal: enabling Europe to govern complex, AI-augmented software environments with the security and institutional intelligence the next decade demands.
Deep tech is patient work.
Seerene
Software portfolio intelligence for large industrial organisations.
Point Cloud Technology
Spatial data infrastructure for government and industry.
Digital Masterpieces
Image and video processing at research-grade quality.
GDT Quantum
Algorithmic trading infrastructure for institutional asset managers.
German Prop Tech
Property analytics and management for complex portfolios.
Sialogic
Operational data intelligence for industrial production.
3D Content Logistics
IoT-based production process analysis and visualization.
German Block Tech
End-to-end infrastructure for digital financial markets.
…