Best Practices

Privata works best when projects, source documents, collections, instructions, and AI workflows are kept aligned. The following best practices will help you produce stronger responses, generate more useful proposals, and keep your work organized across the full platform.

INFO

SkyPath Privata is designed for confidential AI-assisted work. The platform uses stateless AI workflows together with collection-based context, so output quality depends heavily on the relevance of the source documents, collections, and instructions you provide.

1. Organize One Project Per Opportunity

• Create one project for one bid, contract, customer response, or proposal effort.
• Keep all source documents for the same activity inside that project.
• Use clear project names so the workspace is easy to recognize later.
• Add a description when it will help distinguish similar opportunities.
• Update project status as work progresses so the workspace reflects where the effort stands.

Projects are intended to keep the source material, supporting knowledge, instructions, and generated content for one sales activity in one place.

2. Keep Source Documents and Collections Separate

• Use source documents for the RFI or RFP files that need to be answered.
• Use collections for the supporting knowledge that helps the AI tools produce stronger output.
• Do not treat the source document and the knowledgebase as interchangeable.
• Attach only the collections that are relevant to the current project.

This separation makes it easier for Privata to use the correct context during answer generation or proposal drafting.

3. Build Focused Collections

• Group documents by product, service line, solution type, or response theme.
• Use descriptive collection names and add descriptions whenever possible.
• Include content such as product guides, pricing tables, statements of work, company background documents, previous RFIs, previous proposals, and technical reference material.
• Avoid placing your entire library into one collection unless there is a clear reason to do so.
• Review collection contents regularly so outdated or conflicting material does not weaken AI output.

TIP

Smaller, focused collections usually produce better results than one large general-purpose collection. If you are unsure whether a document belongs in a collection, ask whether it would help answer the current opportunity accurately.

4. Use Clean, Relevant Source Material

• Upload the final or near-final version of the source document whenever possible.
• Use supported formats such as PDF, DOC, DOCX, XLS, XLSX, CSV, PPT, PPTX, JPG, JPEG, PNG, TIFF, BMP, TXT, and MD where applicable.
• When adding websites, choose stable and authoritative pages such as white papers, user guides, and case studies.
• Make sure uploaded documents finish indexing before relying on them for AI generation.
• Remove duplicate, obsolete, or low-value material from collections to reduce noisy results.

Clear, relevant inputs give the platform a better foundation for both Q&A and proposal workflows.

5. Choose the Right Response Method

• Use the Q&A Method when the document contains individual questions, requirements, or response items that should each receive a separate answer.
• Use the Proposal Method when the document should be evaluated as a whole and answered with one narrative solution.
• If the final deliverable should read like a list of question-and-answer pairs, use Q&A.
• If the final deliverable should read like a standalone solution pitch, use Proposal.

Choosing the correct workflow at the start reduces rework and helps Privata generate the kind of output you actually need.

6. Review Instructions Before Generating Content

• Use the instruction extraction option when source documents contain formatting rules, word limits, or customer-specific directives.
• Review extracted instructions before generating content.
• Remove instructions that are incorrect, duplicated, or no longer relevant.
• Add manual instructions when you need to guide tone, structure, compliance language, or specific response constraints.
• Keep instructions clear and non-conflicting.

Instructions are one of the most important controls for improving output quality across both Q&A and proposal workflows.

7. Edit Prompts Deliberately

• Use Edit Prompt when you need to change tone, structure, detail level, or creativity.
• Keep prompt changes specific and intentional instead of adding too many competing directions.
• In the Q&A workflow, remember the prompt is centered on the extracted question plus instructions and collection context.
• In the Proposal workflow, remember the prompt applies to the source content as a whole and is intended to shape the final narrative response.
• Recheck output carefully after changing prompts in ways that increase creativity.

Prompt editing works best as a targeted adjustment, not as a replacement for strong collections and clear instructions.

8. Review AI Output Before Final Delivery

• Treat generated content as a draft that should be reviewed before delivery.
• Use the WYSIWYG editor to revise wording, apply formatting, and improve clarity.
• Save changes as you go, especially on larger RFIs and proposals.
• In the Q&A workflow, review low-confidence answers first.
• Use required and optional indicators to prioritize review, even though Generate All will answer every item.
• For proposals, check that the final narrative reflects the actual solution, services, integrations, and deliverables you intend to offer.

Careful review is especially important when responding to sensitive opportunities or documents with strict compliance requirements.

9. Download and Merge with Care

• Use Word, PDF, or text when you need a standalone deliverable.
• Use Excel or CSV when you need tabular output from the Q&A workflow.
• Use Download Merge with Original only when the source document structure clearly shows where answers should be inserted.
• For spreadsheets, make the answer location obvious before upload whenever possible. A column named Answers is better than ambiguous columns such as Description, Comments, or Vendor.

WARNING

The merge feature in the Q&A workflow may not always place answers exactly where you expect if the original document does not clearly indicate where responses belong. Review merged files carefully before sending them to a customer or stakeholder.

10. Protect Work from Avoidable Loss

• Double-check before deleting a project or collection.
• Keep important source documents and response exports in your normal document management process.
• Confirm that a project or collection is no longer needed before removing it.

DANGER

Deleting a collection permanently removes its indexed data and uploaded documents. Deleting a project removes its source documents and generated responses. These actions cannot be undone.

End-to-End Workflow Checklist

Details

• Create and verify your account, including SMS verification.
• Create one project for one opportunity.
• Upload the correct source documents for that effort.
• Choose the right response method: Q&A or Proposal.
• Review and clean up extracted instructions.
• Add focused collections with relevant supporting documents and websites.
• Wait for indexing to complete before generating content.
• Generate answers or the proposal draft.
• Edit the output in the WYSIWYG editor and save your work.
• Review low-confidence Q&A items and inspect any merged output carefully.
• Download the final response in the format best suited to delivery.

Related Guidance

For more detailed workflow guidance, continue to the Quick Start Guide, Projects, Building Your Knowledgebase and Collections, Working with Q&A Responses, and Generating Proposals pages.