See it. Understand it.
Act on it.
Build workflows that read images, documents and video, check them against your business data, flag exceptions and take action. Visually, with no code.
From a photo to a verified business action.
OCR alone only gives you text. RunOCR is for work where visual information has to be understood, checked against business context and turned into actions, with people involved where it matters.
Capture photos, scans, PDFs, video frames and web pages from uploads, APIs, webhooks or schedules.
Vision AI extracts typed fields, products, codes, objects and visual changes.
Match against your records, compare with tolerances, check confidence and catch duplicates.
Create and update records, call APIs, send alerts, open exceptions or ask a person.
One set of steps for almost any visual workflow.
No industry templates you have to bend to fit. Combine 50+ configurable steps into your own process on one canvas.
- Image or video uploaded
- Website captured
- Scheduled execution
- API request or webhook
- New record
- Extract text (OCR)
- Structured field extraction
- Detect objects & products
- Barcodes and QR codes
- Compare images & detect changes
- Search your database
- Find or match records
- Fuzzy-match people
- Query reference datasets
- Call external APIs
- If / Else and Switch
- Compare with tolerance
- Threshold & confidence checks
- Duplicate detection
- Required field rules
- Classify information
- Infer attributes
- Summarize findings
- Detect anomalies
- Map to your schema
- Create & update records
- Send email & notifications
- Call APIs & webhooks
- Generate reports
- Create exceptions
- Request approval
- Manual verification
- Assign to a teammate
- Request more evidence
- Escalate exceptions
Save any group of steps (for example “Verify Employee Identity”) and drop it into other workflows.
Same building blocks. Your rules.
All four workflows use the same steps. Only the configuration, reference data and business rules change. Each one is a ready-made template in the app.
A canvas built for real operations
Drag steps onto the canvas, connect their outcomes and configure each one in the side panel. Insert any earlier step's output with a variable picker. Switch to Debug and watch a test run move through the graph.
Your own component library
Package proven logic once and reuse it everywhere:
AI that proposes. Rules that decide.
RunOCR treats AI as a source of typed evidence. Business actions only happen through rules you can read, or through people you trust.
Every vision and AI step returns fields you define. Required fields that come back empty fail the step, so nothing downstream acts on a guess.
AI never triggers actions directly. Conditions are plain rules over typed values, so you can see exactly why each branch was taken.
Pause any run for approval, correction or more evidence. Reviewers get a secure link and don't need an account.
Publishing freezes a version. Running workflows never change under your feet, and you can roll back in one click.
Debug mode shows every step's input, output, port and timing on the canvas while a test run executes.
Each step has its own retry and timeout settings, and can route failures to an error branch instead of stopping.
API tokens are encrypted at rest and only added to outgoing requests. They are never shown again after you save them.
Every run keeps its inputs, decisions, reviewer actions and outputs, so you can answer how a record was created.
Start runs from your own systems.
Every published workflow can be started with a REST call or a secret webhook URL. Send JSON or files, then wait for the typed result or receive it through a signed webhook.
- API keys per account
- Multipart file uploads
- Synchronous wait (up to 60s) or polling
- HMAC-signed outbound webhooks
- Records API for reference data
# Start a run with a photo and wait for the result
curl -X POST "https://api.runocr.com/v1/workflows/42/runs?wait=30" \
-H "Authorization: Bearer rk_live_…" \
-F "file=@package.jpg" \
-F "dock=Rotterdam-3"
# → 201 Created
{
"status": "succeeded",
"output": {
"label": { "tracking_number": "1Z999AA10123456784",
"weight_kg": 12.9, "confidence": 0.99 },
"weight": { "same": true, "diff_pct": 3.2 }
}
}Build your first visual workflow today.
Start from a template, upload a real photo and see each step's result in minutes.