Cai Disclaimer

Cai is the AI functionality within Crowdtech Insights, designed to support researchers across different parts of the research process. For example, Cai helps categorise open-ended responses, bringing structure to large volumes of textual data.

Cai supports, the researcher decides

Cai is meant to save researchers time and support them throughout different steps of the research process, giving them more time for analysis, interpretation and drawing conclusions. Cai does not replace the role of the researcher.

Use Cai as a supporting tool within your own research workflow and combine its output with your own expertise, context and judgement.

AI output may contain errors

Although Cai is developed with reliability and quality as key principles, AI output may be incomplete, inaccurate or not optimal. AI models work based on probabilities and may, for example, assign a response to a less appropriate category, misinterpret a response or miss relevant nuances.

Review the output

We recommend always checking Cai’s output, at least on a sample basis, before using it for further analysis, reporting or decision-making. For open-ended response categorisation, this could mean checking whether the suggested categories are appropriate and whether responses have been assigned to the correct categories.

Privacy and data processing

Crowdtech takes the security and confidentiality of research data seriously. Respondent data never reaches the AI model in an identifiable form. Personal data is automatically detected, where possible, and replaced with [redacted] before the data reaches the AI model.

Cai uses European AI technology. For more information about how Crowdtech processes personal data and research data, please visit our Trust Center.

Continuous development

AI technology is continuously evolving. We will continue to improve Cai in terms of quality, reliability and ease of use. As a result, the functionality and results of Cai may change over time.