Skip to content
TheChatArchiveThe Chat Archive
NVIDIAUnspecified NVIDIA model (HelpSteer2)· NVIDIANVIDIA NVIDIA HelpSteer2

we are using angular with AWS for our care platform. Using OpenAI GPT to analyze patient…

  1. User

    we are using angular with AWS for our care platform. Using OpenAI GPT to analyze patient and care manager converation. Instead of passing the API so it can viewed in the payload, what are better ways to pass this information so the prompt and payloads are protected

  2. NVIDIA Assistant

    There are several ways to protect the prompt and payload when using OpenAI GPT for conversation analysis in an Angular-based care platform on AWS:

    1. Use an API gateway: You can use an API gateway to securely route requests to the OpenAI GPT model. The API gateway can authenticate and authorize access to the model, and can encrypt the data in transit using SSL/TLS.

    2. Use AWS Lambda: You can use AWS Lambda to securely invoke the OpenAI GPT model. Lambda can be configured to require authentication and authorization, and can encrypt the data in transit using SSL/TLS.

    3. Use AWS Secrets Manager: You can use AWS Secrets Manager to securely store the credentials and other sensitive information required to access the OpenAI GPT model. Secrets Manager can be configured to require authentication and authorization, and can encrypt the data using AWS Key Management Service (KMS).

    4. Use AWS Identity and Access Management (IAM): You can use AWS IAM to securely manage access to the OpenAI GPT model. IAM can be used to create and manage user and group permissions, and can be used to encrypt the data using AWS KMS.

    5. Use AWS CloudFront: You can use AWS CloudFront to securely distribute the OpenAI GPT model. CloudFront can be configured to require authentication and authorization, and can encrypt the data in transit using SSL/TLS.

    By using one or more of these methods, you can help ensure that the prompt and payload are protected when using OpenAI GPT for conversation analysis in your Angular-based care platform on AWS.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.