Skip to main content
This 5-step guide walks you through your first conversation in Omnifact. You will sign in, start a chat, send a prompt, review how the Privacy Filter processes sensitive data, and explore AI models and attachments.
1

Sign in to Omnifact

Open your web browser and navigate to https://chat.omnifact.ai. Enter your account credentials or sign in using Single Sign-On (SSO) if configured by your organization.
Omnifact Login Screen
2

Start a new conversation

Once signed in, Omnifact opens directly into a new chat. To start a fresh conversation at any time, click New chat at the top of the sidebar or press Cmd/Ctrl + J.
New Chat action at the top of the sidebar
3

Send your first prompt

In the chat input at the bottom of the screen, enter your question or task. Start with a direct prompt:

Draft an email

“Draft a concise email to my project team summarizing our weekly progress and outlining three next steps.”

Summarize or structure

“List the main differences between relational databases and document stores in a comparative table.”
Press Enter or click the send button to submit your message. Use Shift+Enter to insert line breaks.
Typing your first prompt in the chat input
4

Review the agent's response

Omnifact generates responses in real time using word-by-word streaming. Before your input reaches the external AI model, Omnifact’s scans the text:
How the Privacy Filter works: Omnifact detects and replaces supported sensitive information (such as personal names, email addresses, phone numbers, and financial identifiers) before the prompt is sent to the selected model provider. When the model responds, the original values are automatically restored in the text.
The Privacy Filter protects supported PII and financial records by default. It does not inspect unsupported custom data types or external links. See How the Privacy Filter Works for supported entity types and Click to Reveal behavior.
Omnifact AI Response
5

Select an AI model or attach files

Customize your conversation using the controls inside the chat input:
  • Choose an AI model: Click the model selector in the chat input to switch between available models from providers like OpenAI, Anthropic, Google, Mistral AI, or Azure OpenAI.
  • Attach files: Click the + (Plus) attachment icon in the chat input to upload files for analysis.
    • File limits: Up to 10 attachments per message. Images can be up to 20 MB; audio files up to 25 MB.
    • Supported formats: PDF, Word, Excel, PowerPoint, OpenDocument, plain text (TXT, CSV, Markdown, JSON), images (PNG, JPEG, WebP, and others), and audio (MP3, WAV, and others).
For comprehensive guidelines on file handling, see Attaching Documents & Images.

​
More Things You Can Do

Once you are comfortable with basic chat, explore these core platform capabilities:

Explore Spaces

Select Spaces in the main navigation to join or create dedicated project workspaces. Spaces feature specialized chat instructions, assigned team members, and a dedicated Knowledge Base.

Add Your Role and Response Preferences

Go to Account Settings > Personalization and complete What do you want Omnifact to know about you? and How do you want Omnifact to respond back to you? The agent incorporates these preferences into its answers.

Connect Cloud Storage

If enabled by your organization, link Google Drive, OneDrive, or SharePoint in Account Settings > Connections to import documents directly into chats and Spaces without manual file uploads.

Administer Your Organization

If you are an organization administrator, use Team Settings to configure authentication, data privacy, user groups, and AI model access.

​
Next Steps

Explore detailed guides to deepen your workflow: