Quantra Documentation

Platform User Guide

This guide covers the day-to-day use of the Quantra platform, from logging in and creating a project, through designing and running a pipeline, to reviewing the results. It is written for end users. Technical and developer information is in later sections.

Logging In and the Dashboard

After signing in, you are taken to the dashboard. The dashboard is your starting point and shows every project you have access to.

  • My Projects — Projects you own. Each card shows the project name, description, creation date, and any schedule that is set.
  • Shared With Me — Projects another user has shared with you. The card includes a badge showing the permission you have been granted (Read, Execute, or Write).
  • Running indicator — A small "Running" pill appears on a project card while one of its pipelines is executing. The dashboard refreshes automatically every few seconds, so the indicator updates without needing to reload the page.
  • Schedule indicator — A small clock pill appears on a project card if the project has a one-time or recurring schedule attached.

The dashboard sidebar contains four actions: Create (start a new project), About, Settings (your account settings), and Logout.

Working with Projects

A project is the top-level organisational unit in Quantra. Each project contains a single canvas — one pipeline graph — along with its execution history, schedules, sharing settings, and audit log. Projects are also the boundary for access control: another user can only see a project if you share it with them.

Creating a Project

Click Create in the dashboard sidebar, then enter a name and a short description for the project. Choose a descriptive name so you can find the project later (for example, "Q4 Invoice Review" or "Customer Onboarding Documents"). After creating the project, you are taken straight to the canvas.

Editing or Deleting a Project

Only the project owner can rename, edit the description of, or delete a project. Deleting a project permanently removes its canvas, execution history, schedules, and audit log. This action cannot be undone.

The Canvas

The canvas is the visual workspace where you design a pipeline. Opening a project takes you straight into its canvas.

Canvas Layout

  • Left toolbar — The vertical toolbar on the left contains the main actions: Run, Schedule, Share (owner only), Audit logs, Go Back (returns to the dashboard), and Logout.
  • Add Tool button — A circular Add tool button at the bottom-right of the canvas opens the node selection dialog.
  • Workspace — The rest of the screen is the canvas itself. Nodes appear here as cards that you can drag around. The workspace supports panning and zooming with the mouse.

Auto-Save

The canvas saves automatically. Any change you make — adding, moving, configuring, or removing a node, or drawing or removing an edge — is saved as soon as you make it. There is no separate "Save" button.

Adding Nodes

Click the Add tool button at the bottom-right of the canvas to open the node selection dialog. The dialog groups the available items into three categories that match the role they play in a pipeline:

  • Datasources — Bring data into the pipeline. Examples include network file shares, mailboxes, document libraries, databases, and cloud storage. A pipeline normally starts with one or more datasource nodes.
  • Tools — Process, transform, analyse, or enrich data as it flows through the pipeline. Examples include Optical Character Recognition (OCR) for printed text, Handwritten Text Recognition (HTR) for handwritten text, Personally Identifiable Information (PII) detection, redaction, summarisation, hashing, and archiving.
  • Workbenches — Interactive review interfaces that let a person inspect, annotate, redact, or approve documents at a specific point in the pipeline.

Use the search box at the top of the dialog to filter the list. Selecting an item adds it to the canvas. You can then drag the node to wherever you want it.

Configuring a Node

Click a node to open its configuration window. The fields you see depend on the node type — a datasource asks for connection details, a tool exposes its processing options, and a workbench has its own settings. Required fields are clearly marked. Closing the window saves your settings.

Node Status

While a pipeline is running, and after it has finished, each node shows its current status with a coloured badge:

  • Pending (yellow) — The node is queued or currently working.
  • Success (green) — The node finished successfully.
  • Partial (orange) — The node finished, but some items did not process successfully.
  • Failed (red) — The node encountered an error. The error message is shown when you open the node.
  • Needs Authentication — The node could not run because credentials are missing or have expired. Open the node and re-authenticate to continue.
  • Cancelled — The pipeline was cancelled before the node could finish.

Connecting Nodes

Edges define how data moves between nodes. To create an edge, drag from a node's output port to another node's input port. A line appears connecting the two nodes.

Edge Types

There are four edge types. The type controls what happens between the two connected nodes:

Flow Edge

The standard edge. Data produced by the upstream node is passed automatically to the downstream node when the upstream node finishes. Most connections in a pipeline are flow edges.

Interactive Edge

An interactive edge pauses the pipeline at the downstream node and waits for a person to take action — typically reviewing a document in a workbench. The pipeline only continues once the reviewer is done.

Reference Edge

A reference edge passes a pointer to the data rather than a copy of the data itself. This is used when a downstream node needs to know about upstream data — for example, to cite the original document — without needing the full content again.

Helper Edge

A helper edge attaches a helper to another node. Helpers are nodes that provide an additional capability to the node they are attached to, such as an Artificial Intelligence (AI) helper that a summarisation tool can call. The helper itself does not appear as a separate stage in the pipeline.

Running a Pipeline

Click Run in the left toolbar to start the pipeline. Before the run starts, the canvas is checked for problems such as missing required configuration, disconnected nodes, or loops. If anything is wrong, the offending nodes are highlighted and a message tells you what to fix.

While the Pipeline Runs

A run progresses through these stages: QueuedRunning → one of Success, Partial, Failed, Needs Authentication, or Cancelled. The dialog updates automatically as the pipeline progresses, showing how many nodes have completed and the overall progress. You can leave the dialog and return to the canvas at any time; the dashboard will continue to show a "Running" indicator on the project card until the run finishes.

Re-Running

To run the pipeline again, click Run again. Each run is recorded as a separate entry in the execution history.

Execution History

Every pipeline run is recorded. You can review the most recent runs of a project, with the following information for each:

  • Status — Success, Partial, Failed, Cancelled, or Needs Authentication.
  • Nodes executed — How many nodes ran during this execution.
  • Timestamps — When the run was queued, when it started, and when it finished.
  • Duration — How long the run took from start to finish.
  • Error message — For failed runs, the reason the run did not complete.

Scheduling Runs

Click Schedule in the left toolbar to set up an automatic run. Two scheduling modes are supported:

  • One-time schedule — Pick a specific date and time. The pipeline runs once at that time and the schedule is removed afterwards.
  • Recurring schedule — Choose hourly, daily, weekly, or monthly. The pipeline runs automatically each time the schedule comes around.

A scheduled project shows a clock pill on its dashboard card so you can see at a glance which projects have automatic runs configured. Schedules can be changed or removed from the same Schedule dialog.

Sharing a Project

Projects can be shared with other registered Quantra users. Only the project owner can manage sharing. Click Share in the left toolbar to open the sharing dialog, enter the email address of the user you want to share with, and choose a permission level.

Permission Level View Canvas Run Pipeline Edit Canvas Delete Project Manage Sharing
Read Yes No No No No
Execute Yes Yes No No No
Write Yes Yes Yes No No

Permissions can be changed or revoked at any time from the same dialog. Only the project owner can delete a project or share it further.

Audit Log

Every meaningful action on a project is recorded in its audit log: edits to the canvas, pipeline runs, sharing changes, and configuration changes. Click Audit logs in the left toolbar to open the log. Each entry shows who made the change, what action they took, and when it happened. The audit log is a useful starting point if you need to understand how a project came to be in its current state, or to satisfy a compliance question.

Reviewing Results

How you review results depends on the type of node:

  • Workbench nodes — Open the node's interactive review interface (see Workbenches below). This is the main way to review processed documents.
  • Tools that produce files — Some tools (for example, the redactor and the archiver) produce output files such as redacted documents or compressed archives. These files are made available through the tool's own interface or through a workbench downstream of the tool.
  • Whole-execution results — You can also download a single results file for the whole run, containing the structured output produced by every node in the pipeline. This is useful for keeping a record of what a particular run produced.

Workbenches

A workbench is an interactive review interface that lets a person inspect, annotate, redact, or approve documents at a specific point in a pipeline. Common review actions include navigating through the pages of a document, drawing redaction boxes, adding notes, and approving or rejecting items.

Opening a Workbench

Click the Open Workbench button on a workbench node to open its interface. A workbench can also be anchored to a project: an anchored workbench opens automatically when you load the project, which is useful when most of your work in the project is review work in that workbench.

For a detailed description of each workbench, see the Workbenches section.

Account Settings and Multi-Factor Authentication

Open Settings from the dashboard sidebar to manage your own account.

  • Profile — Update your name, email, and password.
  • Multi-Factor Authentication (MFA) — Enable Time-based One-Time Password (TOTP) authentication using an authenticator app on your phone. When you enable MFA, you are also given a set of recovery codes; keep them somewhere safe so you can sign in if you lose access to your authenticator app.

Your administrator may require MFA for all accounts. If MFA is required and you have not yet set it up, you will be prompted to enrol the next time you sign in.