Which plans have access to AI?
Insights and AI are accessible to Team, Enterprise, and Results & Analytics plans. Results & Analytics can use AI to apply to existing studies.
How can I set my organization's settings preferences for AI?
The choice is always yours whether to use AI for Insight generation. As an admin, you can turn on or off AI by default for your organization at any time. To turn AI on or off for your organization users, go to Workspace Settings > Features > Edit to change your preferences.
If AI is turned off by default by your organization, the Generate button will be disabled for your users and cannot be used to generate Insights using AI.
If you are an admin of a multi-team organization, you can:
Request that AI be turned on for specific Workspaces. Please get in touch with support or your customer success manager to submit a request.
Navigate to your Admin Team Workspace Settings > Features > Edit to change your preferences. Please note that the settings selected will apply to all Workspaces at your organization.
As an alternative, you can request for AI for Insights to be removed completely for your organization. Please get in touch with Optimal to have AI removed from your organization settings.
What tech does Optimal use for AI generation?
Optimal uses AWS Amazon Bedrock, which is the fully managed service that makes large language models (LLMs) from Amazon and leading AI startups available through an API.
Learn more in our How AI generated Insights work in Optimal help article.
How does AI in Optimal create Insights?
We leverage a two-step method to bring this feature to life.
Turning observations into numbers (vectorization): We use advanced generative AI models to convert observations into lists of numbers. This process, called vectorization, captures the key features and meaning of each observation in a numerical format known as a vector.
Clustering for Insights: We then use clustering techniques to group these vectors based on how similar they are. Observations with similar number lists are grouped together, while those with very different numbers are kept separate. This grouping helps us spot important patterns and build Insights from the data.
What privacy and security measures are in place for AI at Optimal?
Amazon Bedrock meets industry-leading standards for compliance, including: ISO, SOC, CSA STAR Level 2, GDPR, and HIPAA eligible. Learn more about Amazon Bedrock.
Amazon Bedrock is a FedRAMP High authorized service in the AWS GovCloud (US-West) Region. FedRAMP High authorized with the highest level of security certification provided under the Federal Risk and Authorization Management Program (FedRAMP) for cloud services used by U.S. federal agencies.
Amazon Bedrock uses encryption to protect data at rest and data in transit.
How does AI in Optimal handle data?
The AI Insights capabilities in Qualitative Insights are separated by customer and the data you submit to AI Insights will not be used to train other AI models.
Limited Use: You control the data you submit and it will only be used within the scope of your chosen study.
No External Model Training: Our AI model does not use your data to train other models.
Ownership: You own all Customer Data outputs generated by our AI model.
Will model training occur at the customer level?
No. Customer data will not be used to train a model available to that specific customer.
How can I set my preferences for AI?
Can I edit AI-generated content?
Yes, you can edit anything AI-generated, including the title, summary/description, and associated observations.
Can AI work with languages other than English?
Our AI-powered Insights are designed to support languages beyond English. The AI Insights functionality works to detect the main language of your observations and generate titles and descriptions in that language.
Important Note: While the AI strives to accurately match the language of your observations, occasional inconsistencies may occur. Your feedback is invaluable for helping us improve this feature, so please share your experience through our feedback and feature request form.