Working with Q&A Responses

When uploading a source document from a project or through the New Project Wizard, you will be asked whether you want to use the Q&A Method or the Proposal Method. This page explains how the Q&A Method works.

The Q&A response workflow is designed for source documents that should be broken into individual questions and answered one at a time. This method works well for RFIs and similar documents where each requirement, question, feature request, or service description should receive its own response.

Choosing the Q&A Method

The Q&A Method is intended for documents that contain separate items requiring individual answers. Privata takes the source document and breaks it into individual sections that represent response items.

These items can include:

• Direct questions
• Requirements
• Descriptions of a feature or functionality
• Descriptions of a service that need a response

The Q&A Method supports source documents such as:

• PDF
• DOC and DOCX
• XLS and XLSX

If your document needs a single consolidated response instead of separate answers, use the Proposal Method instead.

How Question Extraction Works

When you upload a source document using the Q&A Method, Privata uses AI logic to identify what parts of the file represent questions or response items. The system automatically distinguishes likely questions from supporting descriptions, instructions, and related text that may also appear in the document.

After extraction:

• Questions are collected into a numbered list
• The system keeps track of where each question appeared in the original document
• That position data is later used to help merge answers back into the original file

This process makes it possible to answer each item individually while still preserving the relationship between the extracted question and its original location.

INFO

Question extraction is designed to work across common business document formats, but the content and layout of the source file still affect how clearly response items can be identified.

Opening a Q&A Document

After uploading a document with the Q&A Method from either the Project page or the Project Wizard, you can begin working with the extracted response items.

You can open the document in two main ways:

• Use the View icon to open the document and review the extracted questions
• Use the magic wand icon to automatically begin generating answers

This allows you to either review the document first or move directly into AI-assisted response generation.

Generating Answers

The answer generator includes several features that help you create strong responses for your RFI or questionnaire.

You can generate answers in two ways:

• Generate one answer at a time for individual review
• Use Generate All to process all extracted questions in sequence

When Generate All is used, Privata moves through each question one at a time and creates the best possible response based on the question, your selected collections, any extracted or manual instructions, and the current prompt settings.

Editing Answers in the Rich Text Editor

Each generated answer is placed into a WYSIWYG editor so you can review and modify the content before export.

Inside the editor you can:

• Revise the wording
• Apply rich text formatting
• Add or remove content
• Save your work and return later

This makes it possible to combine AI-generated drafting with manual review and final polishing.

TIP

Save your edits as you work, especially on larger RFIs. Saved answers can be revisited later if you need more time to complete the document.

Editing the Prompt

The Edit Prompt button allows you to modify the prompt sent to the AI system. The prompt typically includes the question, project instructions, and collection-based context.

You can use this feature to add temporary guidance or special response behavior for the current answer generation run.

For example, you could instruct the AI to provide an alternative answer style by adding text such as:

Respond first with the best answer you come up with but also provide a second alternative answer using less concern over factual accuracy and more emphasis on creativity.

Prompt editing is useful when you want to influence tone, structure, detail level, or other response behavior beyond the default instructions.

Confidence Scores and Filtering

Each generated answer includes a confidence score. This score reflects how confident the AI system is in the answer it produced.

The toolbar includes a filtering feature that lets you set a Min Confidence threshold. This helps you focus attention on lower-confidence responses that may need additional review, rewriting, or manual verification.

Confidence filtering is especially useful on long RFIs where some answers are well supported by your collections and others require more user oversight.

Required and Optional Indicators

Some extracted questions may include metadata indicators showing whether a response appears to be required or optional.

These indicators are displayed in the question area as:

• Required with a red outline
• Optional with a blue outline

These visual indicators help prioritize your review, but Generate All will still generate answers for every question regardless of whether the item is marked required or optional.

Downloading Your Responses

Once your answers have been generated and reviewed, you can download the completed response set in a variety of formats.

Available formats include:

• Excel
• CSV
• Word
• PDF
• Plain text

These options make it easier to copy and paste responses into a customer template or work with the output outside of Privata.

Merging Answers Back into the Original Document

One of the download options is Download Merge with Original. This feature attempts to place your generated answers back into the original source document you uploaded.

Privata uses the remembered question positions and AI assistance to determine the best place to insert each answer relative to the original content. In many cases this works well, but the result may not always be perfect.

WARNING

The merge process can be ambiguous when the original document does not clearly show where answers belong. This is common in spreadsheets with columns such as Description, Comments, or Vendor, where there may be no obvious response column.

When possible, make the intended response location clearer before uploading. For example, in an Excel file, a column named Answers gives the system a more obvious target for merged output.

Best Practices for Better Q&A Results

• Use the Q&A Method for documents with separate response items rather than a single narrative response
• Attach only the collections that are relevant to the source document
• Review extracted instructions before generating answers
• Use confidence filtering to identify answers that need closer review
• Edit the prompt when you need a different style or additional output behavior
• If you plan to merge into a spreadsheet, make the answer column easy to identify

What to Do Next

After generating your Q&A responses, the usual next steps are:

• Review each answer in the editor
• Save any manual changes
• Filter low-confidence items for extra attention
• Download the completed responses or merge them back into the original file

For related guidance, continue to the Projects, Building Your Knowledgebase and Collections, and Quick Start Guide pages.