Smart Data Import: Frequently Asked Questions

This feature is currently in beta.

Upload your data in any format – Smart Data Import uses AI to identify variables and get your data ready for graphing, eliminating manual data entry. 

Below are answers to common questions about how AI is applied, what data is processed, and how your privacy is protected.

What is AI used for in the graphing product? 

For the Smart Data Import feature, we use AI (LLMs) to help us identify the variables present in the data, infer the variable types, and infer variable or column names.

What AI models are you using? 

We are currently using Anthropic’s Claude and OpenAI’s 4o and 4o-mini models. We may choose to use other vendors or models in the future, but will only use those that meet the same security and privacy standards.

How much of my dataset gets sent to AI models? 

Currently we send headers (column names) and the first 3 rows of the dataset to help us identify the variables present in the data, infer the variable types, and infer variable or column names. We do not send full datasets through AI models.

Who can view my datasets? 

Only you and anyone you share your graph files with. BioRender’s software processes your data to generate graphs, but BioRender employees will not view your datasets unless you choose to share them when you report an error for troubleshooting purposes. 

Will you use my datasets to train AI? 

Third parties (e.g. OpenAI) will never train their models on your data. We will not train our own AI models on private customer data without explicit approval.

How do you ensure AI does not alter the data or hallucinate? 

The only edits to your spreadsheet the AI might make are to clean up annotations (non-data rows) and clean up or infer column names. It should not change the underlying data points. However, since AI can make mistakes, we always recommend that you verify that your data has been imported correctly.