Seminar in Cincinnati: Deploying GenAI in Market Research with Bayesian Networks and BayesiaLab
Free Workshop in Cincinnati
August 12, 2025, from 2:00 p.m. to 5:00 p.m. (EDT, UTC−04:00)
Indiana Wesleyan University - Cincinnati Education and Conference
Center, 9286 Schulze Drive, West Chester Township, OH 45069
Join us on August 12, 2025, at 2:00 PM EDT in Cincinnati for a free workshop that demonstrates how to unify human domain knowledge, survey data, and GenAI-derived insights into a single Bayesian network model to gain a deeper understanding of consumers.
Conceptual Overview
Bridging the Qual–Quant Divide
Market research is traditionally divided into qualitative and quantitative streams. While both are commonly used, e.g., focus groups for qualitative insights, surveys for quantitative analysis, they are typically treated as parallel processes. Integration, if it happens at all, is left to the intuition and interpretation of the end user. There is rarely a unified knowledge representation that merges both types of inputs.
Is GenAI Just More Qual?
The emergence of GenAI appears to reinforce the qualitative-quantitative separation. Today’s large language models (LLMs) generate narrative responses based on statistical patterns learned from text. This makes them effective at synthesizing open-ended responses or summarizing consumer feedback, but not at analyzing numerical survey data.
GenAI cannot compute even a simple average. Instead of performing the calculation, it might suggest code for doing so in Python or R. In that sense, GenAI is not a computational device but a linguistic one, it mimics answers rather than calculating them.
Bayesian Networks: A True Qual–Quant Integration Framework
Bayesian networks uniquely support the integration of qualitative and quantitative knowledge into a single formal model. Each Bayesian network consists of:
- A qualitative structure (a directed acyclic graph encoding relationships between variables), and
- Quantitative parameters (probabilities or conditional distributions).
These models can be:
- Expert-built, based on domain knowledge;
- Machine-learned from data;
- Or a hybrid, combining both sources.
A new, third source, i.e., GenAI-derived knowledge, will be the main topic of this seminar.
A key advantage is that Bayesian networks natively handle uncertainty, whether from incomplete data, conflicting information, or subjective expert input.
BayesiaLab: Building and Using Integrated Models
BayesiaLab is the leading software platform for constructing, analyzing, and reasoning with Bayesian networks. It enables users to:
- Encode knowledge from multiple sources,
- Perform inference, forecasting, and sensitivity analysis,
- Distinguish between observational and interventional predictions by explicitly modeling causal relationships.
BayesiaLab makes the Bayesian network not just a conceptual framework, but a computational engine for decision support.
Hellixia: Enabling GenAI as a Knowledge Contributor
With the release of BayesiaLab’s Hellixia module, GenAI becomes a powerful knowledge input, not just a qualitative summarizer. Hellixia conducts structured queries to LLMs and translates their responses into Bayesian network components. This allows:
- Generating knowledge graphs, semantic networks, causal networks, etc.;
- Quantification of relationships, by asking GenAI for numeric assessments;
- Comparison of multiple LLM outputs, treating them as expert opinions, and
- Capturing uncertainty arising from diverging assessments.
Hellixia thus overcomes the “qual-only” nature of GenAI by converting its narrative output into structured, quantifiable insights that can be integrated with human expertise and empirical data.
Seminar Topics & Software Demos
- Narratives to Networks: Building structural Bayesian network models from textual consumer feedback using Hellixia;
- Clustering of variables using a Bayesian network machine-learned from survey data with BayesiaLab;
- Multiple-Clustering for latent variable/factor generation utilizing LLM-produced factor names;
- Analysis of key purchase intent drivers and optimization of product characteristics;
- High-dimensional data clustering for consumer segmentation using LLM-generated segment names;
- Target group definition: aligning consumer segments and media channels.
About the Presenter
Stefan Conrady
Stefan Conrady has over 20 years of experience in decision analysis, analytics, market research, and product strategy, having worked with Mercedes-Benz, BMW Group, Rolls-Royce Motor Cars, and Nissan across North America, Europe, and Asia. As Managing Partner of Bayesia USA and Bayesia Singapore, he is widely recognized as a thought leader in applying Bayesian networks to research, analytics, and decision-making. Together with his business partner, Dr. Lionel Jouffe, he co-authored Bayesian Networks & BayesiaLab — A Practical Introduction for Researchers, an influential resource now widely cited in academic literature. With their deep expertise in Bayesian networks for Key Driver Analysis and Optimization, Stefan and Lionel are highly sought-after consultants, advising global leaders such as Procter & Gamble, Coca-Cola, UnitedHealth Group, L’Oréal, the World Bank, and many of the world’s largest market research firms.