Inference Dynamics
Making Complex Reasoning Explicit
Complex challenges are not only difficult because of data—they are difficult because of how different perspectives interact, conflict, and evolve over time.
Inference Dynamics is a research program and analytical method that makes these processes visible.
At its core, it combines:
- qualitative research on expert dialogue and decision-making
- AI-based analysis of language and meaning structures
- system-oriented modeling of interactions and dynamics
Implemented in our system Stratify, this approach enables organizations to move from fragmented discussions to structured, evidence-based decision models.
What We Do
We analyze how stakeholders:
- frame problems
- make implicit assumptions
- generate (often unarticulated) causal inferences
AI allows us to identify patterns in these processes that remain invisible in conventional analysis—revealing hidden structures of reasoning and coordination.
What This Enables
- Clarity in complex decision environments
- Alignment across conflicting perspectives
- Explicit modeling of system dynamics and trade-offs
- Development of actionable strategies and policies
Typical Use Cases
- Strategy development in complex organizations
- Multi-stakeholder governance processes
- Innovation and transformation programs
- Policy design under uncertainty
The Role of AI (Stratify)
Our system Stratify operationalizes Inference Dynamics by:
- structuring qualitative data into dynamic models
- identifying patterns across perspectives
- supporting the design of intervention strategies
This creates a new type of AI-supported research and decision infrastructure.
Partnership
We collaborate with organizations facing:
- high complexity
- multiple stakeholders
- unclear or contested problem structures
Typical partners include corporations, public institutions, and research consortia.
Invitation
If your organization is navigating complex challenges and seeks a structured, AI-supported approach to decision-making, we invite you to connect with us.
Read about our Research
Researching digitalized work arrangements
In this paper, we develop the mathematical basis to study divergent perspectives on complex problems in expert dialogues.
Read the article here:
Rehm, S.-V., Goel, L., & Junglas, I. (2022) Researching digitalized work arrangements: A Laws of Form perspective, Information & Organization, 32(2), June 2022, 100391. DOI: 10.1016/j.infoandorg.2022.100391
Turning Tacit Knowledge into Dynamic Strategies Through
Collective Inferencing
This article introduces collective inferencing, our novel method for making sense of complex organizational dynamics in firms and ecosystems. Drawing on research with the cplace business ecosystem, it shows how generative AI can support the analysis of documents, statements and experts’ tacit insights across stakeholder roles—developers, consultants and managers. By structuring these inputs into a dynamic model, the method helps derive actionable strategies for adaptive governance and decision-making. The article concludes with practical recommendations for applying this approach to platform management, digital transformation and ecosystem innovation.
Read the article here:
Rehm, Sven-Volker (2025) "Turning Tacit Knowledge into Dynamic Strategies Through Collective Inferencing," MIS Quarterly Executive: Vol. 24: Iss. 4, Article 7. Available at AIS.
Observing artifacts: How drawing distinctions creates agency and identity.
Our research presented in this paper explains how our dialogues oscillate around divergent interpretations - and how this allows us to create shared understandings.
Read the article here:
Rehm, S.-V., Goel, L., & Junglas, I. (2023). Observing artifacts: How drawing distinctions creates agency and identity. In Robert D. Galliers and Boyka Simeonova (Eds.), The Cambridge Handbook of Qualitative Digital Research, Cambridge University Press, DOI: 10.1017/9781009106436
Managing Networked Innovation on Digital Infrastructures:
A Cybernetic Method for Collective Sensemaking of Complex Dialogical Problems
In this article, we describe the origins and conception of our Collective Inference Method, and our pilot studies.
Read the article here:
Rehm, S.-V.; Bondel, G. (2021) Managing Networked Innovation on Digital Infrastructures: A Cybernetic Method for Collective Sensemaking of Complex Dialogical Problems. ACM Collective Intelligence Conference 2021 (CI2021), Copenhagen Business School, Denmark [Link]