Pantasso
Plan an evaluation

Concepts

Workspace Objects

How Pantasso structures messages, documents, decisions, projects, and tasks into queryable objects.

Status: Live · Updated 2026-07-12

Workspace Objects

Pantasso organizes imported content into structured workspace objects. These objects are the foundation of cited retrieval, agent context, and workspace health monitoring.

What are workspace objects?

When you connect a source, Pantasso does not store raw data as-is. It structures the content into typed objects that can be searched, cited, and tracked. Each object has:

  • Type: What kind of content it represents (message, document, decision, etc.)
  • Source: Where it came from (which Slack channel, Drive folder, etc.)
  • Content: The text and metadata from the original source.
  • Provenance: When it was imported, when the source was last synced.

Object types

Messages

Imported from Slack channels, Telegram groups, and similar conversation sources. Messages include the text, author, timestamp, and thread context.

Documents

Imported from Google Docs, Google Drive files, Obsidian notes, and uploaded files. Documents include the full text content and metadata.

Decisions

Extracted from messages and documents where a decision was recorded. Decisions capture what was decided, who was involved, and when.

Projects

Structured representations of ongoing work, including scope, status, and related objects.

Tasks

Individual work items connected to projects, decisions, or other workspace objects.

How objects are used

Cited retrieval

When you ask a question, Pantasso searches workspace objects and returns citations pointing to the specific objects that informed the answer.

Agent context

AI assistants with agent profiles receive workspace objects as context when answering company-specific questions. The objects they can access depend on their profile scope.

Workspace health

The workspace dashboard shows object counts, source coverage, and freshness metrics so you can see how complete your workspace is.

Object lifecycle

  1. Import: Content is imported from a connected source and structured into objects.
  2. Index: Objects are indexed for retrieval.
  3. Query: Objects are returned as cited context when relevant to a question.
  4. Update: When sources re-sync, objects are updated to reflect changes.
  5. Archive: Objects from disconnected sources can be archived but remain in workspace history.