Work in the expert tool
We integrate AI agents with the states and workflows of QGIS, geospatial pipelines and CAD systems.
Company
Research, software engineering and learning.
We develop our own specialist software and AI systems and bring this experience to development projects, evaluations and research collaborations. Our Learning Lab makes the underlying methods understandable and accessible.
Our thesis
Expert AI becomes measurable through its work.
AI agents for expert workflows need observable and measurable actions in software: project states, geodata, CAD models, maps, tables and reports.
This is why we develop systems and evaluation methods together. We build technical connections, have agents perform defined tasks, and document outcomes, methods and professional context.
Working principles
What guides our work.
Traceable outcomes
Tasks, data states, artefacts and evaluation methods are documented together.
Continuous development
Domain requirements, implementation, tests and documentation evolve together.
Professional responsibility
Review steps, responsibilities and approvals are integral parts of the workflow.
Our fields
Three fields that complement one another.
Development turns technical requirements into usable systems. Research investigates new forms of AI integration and evaluation. Learning makes the underlying methods accessible.
01 / Development
Expert software for technical workflows.
Our development portfolio comprises expert software and AI systems for a range of technical workflows. SpatialAgents, the Geospatial API for AI agents, Climate Risk and Vulnerability Assessment (CRVA) built on that framework, and Düngewende represent a selection of our development work.
02 / Research
Study AI agents inside expert tools.
Our research programme combines the methodical evaluation of AI agents with the study of new forms of integration in technical software. SpatialAgents benchmarks, CAD-Agents and SpatialApp represent a selection of our research in GIS and CAD.
03 / Learning
Make numerical methods accessible through self-study.
The Holistech Learning Lab combines progressive derivations, visualisations and executable experiments. Its first books cover FEM, electrodynamics and biomechanical modelling of bone remodelling.
Development and research
Collaboration on demanding specialist systems.
We support projects from the domain question through architecture and implementation to methodical evaluation and documentation.
01 / Development
Develop specialist software and AI integration.
We design and implement specialist applications, agent-oriented interfaces, geospatial processing and the integration of AI agents into GIS, QGIS and CAD workflows.
Potential contributionsSystem architecture, APIs, data pipelines, GIS integration, AI agent systems, testing and technical visualisation.
02 / Research
Study and evaluate AI systems methodically.
We develop research designs, task and benchmark formats, and artefact-based evaluation methods for AI agents in specialist software.
Potential contributionsFeasibility studies, method development, benchmark design, repeatable evaluation, research software and transfer into technical systems.
Learning in the Learning Lab
Approaching complex methods from several perspectives.
Our interactive books are designed for self-study and also address students with little prior mathematical experience or a background in fields such as biology.
Using the finite element method as an example, we progressively connect observable phenomena, geometric ideas, mathematics, algorithms, executable code and domain applications. Learners can build on their existing knowledge and develop new connections at their own pace.
Numerical methods help investigate real systems, compare alternatives and improve solutions. AI supports the construction, execution and analysis of models. People contribute the decisive questions, domain judgement and diversity of ideas.
FEM, electrodynamics and biomechanics form the starting point. Further interactive learning aids and numerical methods will extend the Learning Lab.
Legal operator
Institut für holistische Technologieforschung GmbH
- Registered office
- Dahlgrundsweg 5
38312 Börßum
Germany - Represented by
- Sören Gebbert
- Register
- Local Court of Braunschweig
HRB 208486 - Contact
- info@holistech.ai
Archive
Selected earlier work.
Dated archive articles document earlier institute publications. The Cyberdeck self-build archive provides two construction guides as Google presentations.