Basic Chat Features

EmpowerGPT provides a set of intuitive chat controls that allow you to select AI models, refine prompts, and interact with responses efficiently. These features help users iterate quickly, compare outputs, and extract maximum value from AI-powered conversations.

Prompt Functionalities

When you hover over a prompt you’ve submitted, available actions appear.

Edit Prompt

If you are not satisfied with a response, you can edit your original prompt by clicking the Edit (pen) icon. After making changes, click Save to resend the updated prompt.

Iterating on prompts is a normal and effective way to work with AI models. Adding more context, clarifying intent, or refining instructions often leads to significantly improved responses.

---

Response Functionalities

Hovering over an AI-generated response reveals several interaction options.

Narrate Response

Click the Play icon to listen to the response. This feature is useful for hands-free consumption or reviewing longer outputs.

Copy Response

Click the Copy icon to copy the response to your clipboard. You can then paste the content into documents, emails, presentations, or other tools.

Regenerate Response

If a response does not meet your expectations, click the Regenerate (circular arrow) icon to generate a new response.

AI models are non-deterministic, meaning the same prompt can produce different results each time.

You may also select a different AI model before regenerating to further influence the quality and style of the response.

---

Chat Feedback

EmpowerGPT includes a built-in feedback mechanism that allows users to rate AI responses and provide additional context when a response does not meet expectations. This feedback helps continuously improve response quality, relevance, and user experience.

Providing Feedback on Responses

For every AI-generated response, users can provide feedback using the Like or Dislike options.

Like Response

Select the Like option when a response is:

• Helpful and accurate

• Clear and well-structured

• Relevant to the query

Liking a response helps indicate what works well and reinforces positive response patterns.

Dislike Response

Select the Dislike (thumbs down) option when a response is:

• Response cites wrong document.

• Response out of scope of documents.

• Response is harmful/unsafe.

• Response is not helpful.

When a response is disliked, EmpowerGPT prompts the user to provide additional comments.

Providing Additional Comments

If you choose Dislike, you can optionally add comments explaining:

• What was missing or incorrect

• How the response could be improved

• Whether the issue was related to accuracy, depth, tone, or relevance

These comments provide valuable context and help refine future responses and system behavior.