Skip to main content
This tutorial walks you through the core RecordEngine workflow end to end. By the end you will have created a client, uploaded a document, watched the AI extract its data, and approved it for export.
This tutorial assumes RecordEngine is already installed and running. If you haven’t installed it yet, see the Installation Guide first.

Before you start

You need:
  • RecordEngine running and accessible in your browser (e.g. http://localhost:8501 or your server URL)
  • Admin credentials (set during first-time setup)
  • A sample invoice PDF — any invoice will do, even a printed and scanned one

Step 1 — Log in

Open RecordEngine in your browser and log in with your admin credentials. You’ll land on the Dashboard, which shows a summary of all documents in the system. It’s empty for now — that’s expected. RecordEngine Dashboard

Step 2 — Create a contact

Every document in RecordEngine belongs to a contact. A contact represents a client, vendor, or any entity whose documents you’re managing.
  1. Click Contacts in the left sidebar
  2. Click + New Contact
  3. Fill in:
    • Name: Acme Corporation (or any name)
    • Type: Client
    • Email: optional for now
  4. Click Save
The contact is created with two default folders already inside it: Documents and Review. Creating a new contact
You can set a Default Extraction Profile on each contact so that all documents uploaded to that contact automatically use the right profile. For this tutorial, leave it as Default.

Step 3 — Upload a document

  1. Click Upload in the left sidebar
  2. Drag and drop your invoice PDF onto the upload zone, or click to browse
  3. Set:
    • Contact: Acme Corporation
    • Folder: Documents
    • Extraction Profile: Standard Invoice
  4. Click Upload
Document upload form The document appears in the list with status New, then almost immediately changes to Processing.
The first document after a server restart takes longer to process (up to a few minutes) because the AI model loads into GPU memory. Subsequent documents process in seconds.

Step 4 — Watch the AI extract

Once processing completes, the status changes to Needs Review. This means the AI has finished and is waiting for a human to verify the results. Click on the document to open it. You’ll see a three-panel layout: RecordEngine workspace showing three panels The three panels are:
PanelWhat you see
LeftOriginal document preview (PDF render or image)
CentreAI summary, confidence score, and document metadata
RightExtracted fields, line items, actions, and notes
Scroll through the extracted fields on the right. For a Standard Invoice, you’ll see fields like Vendor, Invoice Number, Invoice Date, Due Date, Currency, Subtotal, Tax, and Total Amount — all filled in by the AI.

Understanding the confidence score

Below the AI summary, you’ll see a coloured badge:
BadgeScoreWhat it means
🟢 High90–100All fields extracted cleanly
🔵 Good70–89Minor uncertainty on one or two fields
🟡 Low50–69Several fields uncertain — review carefully
🔴 Poor0–49Significant extraction problems
A clean, well-formatted invoice should score 80 or above. A blurry scan might score 40–60.

Step 5 — Edit a field (optional)

If any field is wrong or missing, you can correct it directly:
  1. Click Edit in the right panel
  2. Click on any field value and type the correct value
  3. Click Save
Changes are saved immediately and logged in the Audit Log.

Step 6 — Approve the document

Once you’ve verified the extracted data:
  1. In the right panel, click Mark as Ready
  2. The status changes to Ready
  3. Click Approve
  4. The status changes to Approved — the document is now locked against further edits
Document status showing Approved
Once a document is Approved it cannot be edited. If you need to make changes, you can upload a new version, or an Admin can revert the status.

Step 7 — Export

Setting a document to Export status triggers the outbound webhook — this is how RecordEngine pushes structured data to your accounting system, CRM, or any other endpoint.
  1. With the document open, click Export (or Webhook Export)
  2. Status changes to Export
  3. If a webhook URL is configured in Settings, RecordEngine sends a POST request with the full structured payload within seconds
Don’t have a webhook set up yet? That’s fine — skip this step for now. See Settings & Configuration to add your webhook URL when you’re ready.

What you just did

In under 10 minutes you completed the full RecordEngine document lifecycle:
Upload → AI Extraction → Confidence Scoring → Review → Approve → Export
Every step is logged in the Audit Log. Every extracted field is searchable via Advanced Search. And every outbound webhook can be routed to Salesforce, QuickBooks, Xero, or any other system.

Next steps