Knowvio logo

Knowvio Docs

Ask. Cite. Recall.

Knowvio Documentation

Everything you need to operate Knowvio confidently

This in-app documentation covers the complete product workflow from ingestion to grounded answers and reusable output capture.

Understand the core Knowvio workflow and product model.

Overview

Knowvio is a visual knowledge workspace where source material, chat reasoning, and final outputs stay connected.

Instead of disconnected chat history, each answer is grounded in selected context and can be traced back to where it came from.

  • Source-to-answer visibility by design
  • Citation-backed responses from linked context
  • Reusable outputs stored as knowledge nodes

Suggested steps

  1. 1.Create or open a canvas for your topic.
  2. 2.Ingest high-signal source material.
  3. 3.Connect sources to chat nodes before asking.
  4. 4.Save useful outputs as reusable text nodes.

Create a canvas and complete your first end-to-end loop.

Quick Start

Start by creating a canvas from the canvases page. Add your first source node, connect it to a chat node, then ask a focused question.

Once you get a useful answer, convert it into a text node so it stays reusable in your workspace.

  • Create canvas -> add sources -> connect to chat
  • Ask targeted prompts against connected material
  • Save high-value output into text nodes

Suggested steps

  1. 1.Open the canvases page and create a new canvas.
  2. 2.Use the source controls to add URLs, YouTube, PDF, image, or text.
  3. 3.Create a chat node and link relevant source nodes to it.
  4. 4.Run your prompt and verify citations in the answer.
  5. 5.Convert strong answers into text nodes for long-term reuse.

Ingest links, files, and text through one consistent flow.

Source Ingestion

Knowvio supports multiple input types including web pages, YouTube links, PDFs, images, and direct text entry.

All ingested content is structured for retrieval so you can query it through connected chat nodes later.

  • YouTube, web, PDF, image, and text support
  • Unified ingestion UI with connected context
  • Ready for semantic retrieval and citations

Suggested steps

  1. 1.Choose the source type from the floating controls.
  2. 2.Provide the link, upload file, or paste raw text.
  3. 3.Wait for extraction/processing to complete.
  4. 4.Review node content and reposition it on the board.
  5. 5.Link source nodes to active chat nodes before asking.

Work visually and keep relationships easy to inspect.

Canvas Workflow

Use the infinite canvas to map source nodes, chat nodes, and output nodes. Connect related items so reasoning paths stay visible.

Use keyboard shortcuts, pan/zoom controls, and fullscreen mode to navigate dense research quickly.

  • Link nodes to preserve reasoning paths
  • Use fullscreen and shortcuts for speed
  • Keep board layout readable as complexity grows

Suggested steps

  1. 1.Arrange source nodes by theme or stream.
  2. 2.Place chat nodes near the source clusters they query.
  3. 3.Use linking mode to connect sources with chat nodes.
  4. 4.Keep final text/output nodes in a dedicated output lane.
  5. 5.Use the nodes panel for fast navigation in large canvases.

Generate grounded answers with verifiable context.

Chat and Citations

Chat nodes answer based on currently connected sources. Citations are attached so outputs remain auditable.

If required context is missing, add or connect more source nodes before re-running your prompt.

  • Answers depend on connected source context
  • Citation references for output verification
  • Iterative prompt + source refinement loop

Suggested steps

  1. 1.Ask one focused question per prompt for cleaner outputs.
  2. 2.Check citations for source coverage and confidence.
  3. 3.If context is missing, connect additional source nodes.
  4. 4.Re-run with refined prompts and narrowed scope.
  5. 5.Save validated output as text nodes for future use.

Use keyboard controls to move faster on complex canvases.

Shortcuts and Controls

Knowvio includes keyboard shortcuts for editing, navigation, and quick actions directly inside the canvas.

Use these consistently to reduce friction while working on larger boards.

  • Undo/redo for rapid iteration
  • Arrow-key panning for precise navigation
  • Fast linking and cancel flows during node mapping

Shortcut reference

Undo

Cmd/Ctrl + Z

Redo

Cmd/Ctrl + Shift + Z or Cmd/Ctrl + Y

Pan canvas

Arrow keys (hold Shift for faster pan)

Toggle fullscreen

F

Add source from clipboard URL

Cmd/Ctrl + V

Close panels / cancel linking / exit fullscreen

Esc

Manage profile and model credentials safely.

Account and Settings

Configure profile details, API keys, and workspace preferences from settings.

API keys added in settings are encrypted before storage, so raw credentials are not persisted in plain text.

Use provider keys that match your preferred model stack and keep credentials up to date.

If no provider API keys are configured, only the default Groq option works with limited model availability.

When keys are missing or invalid, model-specific requests can fail. Keep only active keys configured and remove old ones to avoid confusion.

Profile updates, avatar changes, and key state checks should be reviewed after saving so your workspace behavior matches your current configuration.

  • Profile and workspace preferences
  • Encrypted API key storage
  • Provider key management by model
  • Default Groq-only limited mode without keys
  • Credential health checks and rotation
  • Provider-specific fallback awareness
  • Access hygiene and periodic key cleanup

Suggested steps

  1. 1.Open Settings and confirm your profile details (name, avatar, and basic account info).
  2. 2.Add API keys only for providers you actively use in canvas chats.
  3. 3.Run key validation where available and remove any failed or stale credentials.
  4. 4.Select a model in chat and verify responses are generated successfully with citations.
  5. 5.Rotate keys periodically and update stored credentials before old ones expire.

Need help fast?

Go back to home or open workspace to apply this documentation directly.