Displayr

User wants to use Displayr AI, but not on certain projects because they don't want Combine to create their variable set labels. This is on a tracking study where data and labels are updated over time. In addition to scenarios where users may not want their variable set labels hardcoded for particular projects, certain projects could also have different data security rules and may be prohibited from using openAI on specific projects/data sets.

In order to utilize ChatGPT plug-in in Displayr, a user requires a paid-for ChatGPT license. Allow users to purchase a license and ChatGPT as a 'bundle'.

In the mid term, we are interested in entering prompts to get instant insights. This feature would enable us to quickly obtain relevant information and insights from our data without extensive manual analysis.

I prefer using Displayr's proprietary modeling algorithm for generating themes instead of the open-source AI. However, currently, the only way to disable AI is to turn it off completely, which also disables other AI features. It would be beneficial to have a button that allows users to select between using AI and Displayr's proprietary algorithm when generating themes.

Client would like to add Copilot to Displayr with our current AI offerings.

Sometimes I ask the AI to code it, and then I find my codeframe is crap, and I have to rerun it. It might cost you a lot when I code the whole set of responses again. In another AI coding tool I used in the past, you could tell the AI to code 100 responses first and then when you are happy do the full thing. It might be a good way to save some costs.

The user would like to be able to save their prompt along with the question they are coding. This would be helpful for future reference in tracking studies and for reusing well-crafted prompts in similar future projects.

When users manually assign a response to a theme in Text Categorization, the AI learns from those inputs in real time and automatically applies similar classifications to other, similar responses. Another use case for this feature is in tracking studies. If a user has manually classified responses in earlier waves of a survey, the AI will automatically apply the same classifications to similar responses in newly imported data.

Create a set of predefined prompts based on known use cases (such as sentiment analysis) that users can utilize for text categorization. The common use cases are: https://help.displayr.com/hc/en-us/articles/12508720089359-Optimizing-Text-Categorization-with-Custom-Prompts

When a document's data file is moved in the folder system, the document's file source should automatically update to reflect the new file location. That is, it should be possible to move the file without first copying it, moving it, then updating the document, and only then delete the data file from its original location.