🀝 Collaborate to build better models#

Collaboration is important in AI QA for various reasons:

  • Alignment Between Stakeholders: If your quality criteria are very specific to your business, you will need to engage in some discussions before defining tests, evaluation criteria, or data slices.

  • Accountability Among Stakeholders: Have your tests and issues reviewed by different stakeholders to share responsibility across the team.

With Giskard, you can:

  • Easily invite business experts with domain knowledge to debug your issues.

  • Collect feedback from other team members on your issues.

  • Discuss quality issues with various stakeholders.

  • Integrate your discussions into your favorite communication channels (such as Slack, email, etc.).

Feedback Discussion

➑️ To invite users, you simply need to create a user tab in the Giskard Settings of the hub. Then, you can share your project with the user you just invited.

πŸ’‘ Try it live with our Hugging Face space: here