Analysis at the scale you have it, and every claim checkable.
Tabbi Research is a qualitative analysis practice. Our analysts turn interviews and open-ended responses into findings your team can act on, and every claim we deliver is linked to the responses it rests on, so anyone who doubts it can check.
Led by the researcher who wrote the method.
Andrew Katz, PhD, founded Tabbi Research and leads its analysis. He developed the way we build codes and themes from open-ended responses and published it in peer-reviewed journals, so the method behind your findings is on the record for your reviewers to read.
You work directly with Andrew and the analysts on your project, from the scoping call to the findings you receive. The people who answer your questions are the people doing the analysis.
How the practice works.
An engagement is a service, delivered by people. What you receive is written for the decision in front of you, and built so that it can be checked.
Our analysts do the work.
You bring the question and the data. Our team reads, codes and counts the responses, at whatever scale you have them, and delivers the findings to you. There is nothing for you to set up or run.
Every claim traces to its responses.
Each count in what you receive opens on the responses behind it, and each quote is a respondent’s own words. A point the data doesn’t support is left out, not softened.
Judgment is marked as judgment.
Recommendations are labelled as ours and set beside the evidence they rest on, so your team can weigh the two separately.
A method your reviewers can read for themselves.
The method is set out in three peer-reviewed papers. The Research page has them in full, with two analyses you can explore.
- GATOS: Generative AI-enabled Theme Organization and Structuring for Qualitative ResearchHumanities & Social Sciences Communications, 2026
- Leveraging Generative Text Models and Natural Language Processing to Perform Traditional Thematic Data AnalysisInternational Journal of Qualitative Methods, 2025
- Using Generative Text Models to Create Qualitative Codebooks for Student Evaluations of TeachingInternational Journal of Qualitative Methods, 2024
Bring us the question your data has to answer.
Tell us what you have and the decision it needs to inform. On a scoping call we’ll say what an analysis can and can’t answer before any work starts.