Privata Docs

Appearance

Using Chat ​

Chat gives you a conversational AI experience similar to popular chat tools, but designed for private, confidential work inside Privata. You get a familiar chat interface while keeping the same secure, stateless AI approach that powers the rest of the platform. Open it from Chat in the top navigation.

Chat is useful when you need help with a specific question, a short piece of writing, or a document-related task without setting up a full project and response.

Why Chat Matters ​

Public chat tools may have terms and conditions that do not fully protect your data from being reused by the company that operates the model. Prompts or interactions may be retained or used in ways that are not suitable for sensitive business, healthcare, defense, or compliance-focused work. Chat is intended for teams that need a more controlled way to work with AI.

INFO

Chat uses the same private, stateless AI system that supports the rest of Privata. The model layer does not retain prompt content for training or fine-tuning.

How Privacy Works in Chat ​

Chat is built around two separate ideas:

  • Stateless AI processing: the prompts sent to the AI are not cached, reused for training, or used for fine-tuning.
  • Private conversation history: Privata stores your conversation history in its own application database so you can return to previous chats and continue working.

That history stays inside Privata and is not shared outside the platform. If you remove a conversation or delete your account, the related history is removed as well.

WARNING

Chat is private within Privata, but you should still follow your organization's policies for handling confidential, regulated, or export-controlled information.

The Chat Workspace ​

Chat uses a two-pane layout:

  • A conversation list on the side, which you can search, holds your past chats so you can pick up where you left off.
  • The message composer is where you type. Above it, an Add collections panel lets you attach the collections you want the AI to use. As the panel notes, checked collections are searched for grounded, cited answers.

Using Collections in Chat ​

Attaching collections is what makes Chat powerful for your own work: the AI answers from your Knowledge Base and cites the material it used. Attach or detach collections per conversation.

This is useful for:

  • Testing how accurate a response may be for an RFx-style question
  • Refining a better answer before using it in a formal response
  • Querying your own documents for product, service, or infrastructure details
  • Looking up information from guides, statements of work, case studies, or technical references

For example, if a product installation guide is in a collection, you can ask Chat for that product's minimum system requirements and get an answer grounded in your stored material, with citations.

INFO

Collections belong to your company workspace, so you can chat against the collections your teammates built, not just the ones you created yourself.

TIP

Attach only the collection or collections relevant to your question. Focused context usually produces a better answer than attaching too much unrelated material.

How Chat Differs from the Response Editor ​

Chat is for flexible, conversational work and does not require a source document. If you need structured question extraction, per-answer confidence scores, approved-answer reuse, or a downloadable deliverable, build the work in a project with the Response Editor instead.

Common Uses for Chat ​

  • Answering a one-off customer or partner question
  • Drafting an email related to a bid or opportunity
  • Checking product or service details from your document library
  • Testing how an answer might read before generating a formal response
  • Exploring technical or operational details from your collections

Best Practices for Better Chat Results ​

  • Ask clear, specific questions instead of broad or ambiguous prompts.
  • Attach only the collections relevant to the task.
  • Use Chat to test or refine answers before moving them into formal workflows.
  • Review responses before reusing them in customer-facing documents.
  • Delete conversations that are no longer needed.

Managing Conversation History ​

Conversation history is stored so you can return to previous work and keep context inside the application. Remove a conversation when it is no longer needed.

DANGER

If you delete a conversation, its stored history is fully deleted and can no longer be recovered through the platform.

What to Do Next ​