Use of customer data with AI Computer Vision

Purposes

We process the information used with our services for the following purposes:

• Provide our services
• Machine learning and product enhancement and development

Provide our services

Computer Vision is used to visually identify elements on a computer screen, as further described here. We process images provided at design time to ensure automation resilience by checking for backward compatibility when upgrading the AI model, improving the AI model, and responding promptly to customer support requests. We process telemetry data, including usage data, organization ID, image size in megapixels, inference time, end-to-end latency, geolocation, a list of used screen elements, and AI model version, to ensure smooth platform operation and to optimize the user experience based on observed feedback and usage patterns. We use data analytics to measure and evaluate the performance of our services.

We process account information, such as API keys, to provide customers with access to our services and to secure their accounts.

We may also process data that customers send to us for support or as necessary for troubleshooting issues, in accordance with our Support terms.

Machine learning and product enhancement and development

We process images sent at design time to enhance and improve our products and to train the model, ensuring you receive higher-quality outputs. This enables us to build better services and offer more personalized experiences, such as improved product features.

We only use design-time data for product improvement, so we recommend using non-sensitive, non-personal data when designing your automation flows.

Data storage

Data is stored when using this product as selected by you and in accordance with our data residency page, available at Automation Cloud - Data residency, except for the data used in design time which is centralised either in Europe or USA as described in the data residency page.

Integration with generative AI services

AI Computer Vision does not leverage any generative AI features.

Safeguards

We employ the following safeguards in achieving the purposes described above:

• Training data is deleted together with customer data upon request.
• Data are encrypted.
• The customer can delete the data used with the product by submitting through a data deletion request.
• Other data hygiene measures: role based access control to training data, encryption in transit and at rest.

Data processing limitation

If you would like to limit data processing we recommend using an on-premise server.