RFP Answer Tool Demo
back

Ultra fast collaboration

This is a demonstration of an interaction paradigm for working with AI. The RFP question-answering use case serves as a concrete example to showcase how collaborative Text Editing can enable precie quick edits which ultimately are faster then attempting full automation.

Design Philosophy

This tool is built on several core principles that guide how you interact with AI to create better RFP answers:

Iterative Refinement, Not Perfect First Tries

Instead of expecting the AI to generate perfect answers immediately, this tool encourages experimentation. Every adjustment creates a new version, so you can explore different approaches, compare results, and gradually refine your answers. This iterative process often leads to better outcomes than trying to get everything right in a single prompt.

Granular Control at Every Level

You have control at both the paragraph level and the full-answer level. This granular approach lets you make precise adjustments—emphasize specific keywords, adjust tone in one paragraph while keeping another unchanged, or modify the entire answer's style. Rather than rewriting everything, you can target exactly what needs to change.

Non-Destructive Editing

Every change creates a new version. Your work history is always preserved, so you can undo changes, compare versions, and return to previous states. This removes the fear of "breaking" something—you can experiment freely knowing you can always go back.

Decision Points, Not Just Outputs

The tool presents you with options and decision points rather than just generating final text. You choose which keywords to emphasize, which topics to add, and how to adjust tone and length. This puts you in control of the direction, while the AI handles the implementation. It's a collaboration, not automation.

Transparency and Understanding

You can see where content comes from (source quality scores and document references), understand what adjustments were made (answer history), and get contextual explanations for how the AI interpreted your requests. This transparency helps you make informed decisions and learn what works best for your use case.

Keyword-Driven Refinement

Rather than rewriting entire paragraphs, you work with concepts and keywords. You can emphasize important terms, add detail around specific topics, or remove mentions of others. This approach lets you focus on what matters most—the ideas and concepts—while the AI handles the language and structure.

In essence: This tool treats AI as a collaborative partner that offers options and implements your decisions, rather than a black box that generates final outputs. You maintain control, visibility, and the ability to refine at every step—making it easier to create answers that truly meet your RFP requirements.

Controls

Adjusting Paragraphs

Each paragraph has controls on the right side to refine it:

  • Keyword Controls - Use the icons next to keywords to Emphasize, Add Detail, or Don't Mention specific terms
  • Also Mention This - Add topics or information you want included
  • Suggested Keywords - Check boxes to include AI-suggested relevant topics
  • Custom Instructions - Provide specific guidance for how the paragraph should be modified
  • Word Count Slider - Adjust length from -40 to +40 words
  • Conviction Slider - Modify confidence and assertiveness from -40 to +40
  • Apply Adjustments - Click this button to regenerate the paragraph with your changes

You can also manually edit the text directly, copy paragraphs, delete them, or add new paragraphs between existing ones.

Adjusting the Full Answer

Enable "Full Answer Adjustment Options" in the left sidebar to adjust the entire answer at once:

  • Mention This - Add specific topics or points to the entire answer
  • Custom Instructions - Provide detailed guidance for how the answer should be modified
  • Answer Style - Choose between Standard, Technical, Short Sentences, or Detailed
  • Length Change - Adjust the overall length of the entire answer

Click "Adjust Answer" to regenerate the entire answer with your changes.

Checking Sources

Each paragraph shows a "Source Quality" score (0-100) indicating how well it aligns with your knowledge base. Click "Sources" to see the specific documents and pages that informed the content. Lower scores may need more attention.

Finalizing Answers

When you're satisfied with an answer:

  • Mark as Answered - Indicates you've selected a final answer for the question
  • Mark for Review - Flag answers that need further attention
  • Copy Full Answer - Use the button at the top to copy the complete answer to your clipboard