Using Secure Chat

Secure Chat provides a conversational AI experience similar to other popular LLM chat tools, but it is designed for private and confidential work inside Privata. It gives users a familiar chat interface while applying the same secure, stateless AI approach that powers the rest of the platform.

This makes Secure Chat useful when you need help with a specific question, a short response, or a document-related task without creating a full RFI or RFP workflow.

Why Secure Chat Matters

Public LLM chat tools may have terms and conditions that do not fully protect your data from being reused by the company that operates the model. In some cases, prompts or interactions may be retained or used in ways that are not suitable for sensitive business, healthcare, defense, or compliance-focused work.

Secure Chat is intended for teams that need a more controlled way to work with AI.

INFO

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

How Privacy Works in Secure Chat

Secure Chat is built around two separate concepts:

• Stateless AI processing
• Private conversation history inside the platform

The chat prompts sent to the AI system are not cached, reused for model training, or used for fine-tuning. At the same time, Privata stores conversation history in its own application database so users can return to previous chats and continue working.

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

WARNING

Secure Chat is private within Privata, but users should still follow their own organizational policies for handling confidential, regulated, or export-controlled information.

How Secure Chat Differs from Other AI Workflows

Secure Chat is intended for flexible, conversational work. Unlike the Q&A and Proposal workflows, it does not require a full source document to be uploaded and processed first.

This makes it useful when:

• You need a quick answer to a specific question
• You want help drafting an email or short response for an opportunity
• You want to test how a question might be answered before using it in an RFI workflow
• You want to explore your own documents conversationally

If you need structured question extraction, answer numbering, confidence scoring, or merge-back into the original file, use the Q&A workflow instead. If you need a full narrative solution response, use the Proposal workflow.

Using Collections in Chat

One of the most useful Secure Chat features is the ability to include collections in the chat message. This allows the system to use your Knowledgebase documents as context for the response.

Using collections in chat is useful for several reasons:

• Testing how accurate a response may be for an RFI-style question
• Refining a better answer before using it in a formal response workflow
• 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 included in a collection, you can ask Secure Chat for the minimum system requirements for that product and receive a response grounded in your stored material.

TIP

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

Common Uses for Secure Chat

Secure Chat works well for both formal and informal business tasks. Common use cases include:

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

This is especially useful when a complete RFI or RFP does not exist and you only need help with a specific response.

Best Practices for Better Chat Results

• Ask clear, specific questions instead of broad or ambiguous prompts
• Include only the collections that are relevant to the task at hand
• Use Secure 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

Details

Examples of effective Secure Chat use:

• Ask for a concise answer to a customer email about implementation timelines
• Query a collection for supported deployment environments or system requirements
• Test how a product capability answer will read before using it in an RFI response
• Ask for a short summary of a service offering using only the selected collection context

Managing Conversation History

Conversation history is stored so you can return to previous work, continue a thread, and keep context inside the application.

If a conversation is no longer needed, remove it. If your account is deleted, associated conversation history is removed as well.

DANGER

If you delete a conversation its stored history is fully deleted and no longer recoverable through the platform.

What to Do Next

After using Secure Chat, you may want to move the results into a more structured workflow.

• Use Working with Q&A Responses when you want separate answers for extracted questions
• Use Generating Proposals when you need one complete narrative response
• Use Building Your Knowledgebase and Collections to improve the document context available in chat