FAQs
Key questions about the Secoda product
Last updated
Key questions about the Secoda product
Last updated
Wondering if the Seda AI is the right tool for your workspace considering your data governance and privacy policies?
Ensuring our customers feel safe and comfortable using this tool is a top priority for us. Read on to learn how we are ensuring your data is safe and secure.
The only data sent to the LLM provider is the metadata in your workspace. Metadata is defined as properties and documentation describing your resources, including Owners, Tags, Descriptions, Definitions, Popularity, Resource Names, and more. The data within the resource does not leave the workspace.
No, all metadata is specific to your workspace and is not shared or accessible by anyone who isn't granted access to your workspace. You cannot ask Secoda AI about metadata that is not in your workspace.
According to Anthropic, any data sent over through the API is not used to train their Large Language Model. You can read more about Anthropic's policies .
To improve your user experience with Seda AI, your prompts are saved. You can see your previous prompts in the right hand side panel in the Seda AI page.
Anthropic is GDPR compliant and take data privacy very seriously. They have implemented measures to ensure compliance with GDPR and other data privacy laws, such as providing users with the ability to delete their data and implementing data protection impact assessments. Secoda Inc, which manages Seda is SOC2 complaint and also takes security seriously.
Anthropic hosts their servers hosted strictly in the US, so all metadata that is sent to the LLM provider is sent to the US.
Our AI strictly adheres to schema-based data access permissions. If you don't want certain data in Seda, you can remove it from the sync during the integration step.
In Seda, the Popularity of a resource is a piece of metadata pulled directly from the source. It is defined as the number of queries (for a database/data warehouse) or the number of views (for a data visualization tool) over the previous 24 hours from the last sync with that source. When ranking search results by Popularity, Seda uses a combination of external and internal metrics.
Search results are based on a number of factors. The primary factor is the matching of text in the resource title or description to what has been inputted into the search bar. Popularity of the resource (from external and internal metrics) are also used to order the search results.
You can choose when to run an extraction for each integrations. Schedules for syncs can be set to run on a daily, weekly, or monthly basis. One off manual syncs are available as well, and can only be run by Admins.
Seda prioritizes data security by requiring the creation of dedicated database users with specific permissions and the whitelisting of Seda's IP addresses. This approach ensures that Seda accesses only the necessary metadata without compromising the security of your data sources .