Cradl

Overview of Cradl AI - Extract Data from Receipts with Machine Learning

Cradl AI is a machine learning platform that allows developers to build custom document parsing models to extract data from receipts. The intuitive interface makes it easy to train AI models without coding or ML experience.

How Does It Work?

Cradl uses deep learning technology specifically optimized for document parsing. The models are pre-trained on millions of documents to read text and layout like humans.

To build a custom parser:

• Select fields to extract
• Upload receipt images to train the model
• Click train - the model learns the receipt structure
• Test and deploy the model via API

The parser extracts text, numbers, dates, logos, signatures - any data you need.

Features and Benefits

• Extract any field from receipts without templates or rules
• Handles any receipt layout and language
• Provides confidence score for each prediction
• Scales to millions of documents without manual labeling
• Fast processing with 1-8 seconds per document
• Easy to use visual interface - no coding required
• Deploy via API, SDKs, CLI and integrations

Use Cases and Applications

• Receipt OCR API for accounting and bookkeeping software
• Travel and expense report automation
• Mileage tracking apps
• Extract data for analytics and business intelligence
• Mobile receipt scanning apps
• Automate data entry from paper receipts

Who Is It For?

Cradl is designed for developers looking to build the following types of solutions:

• Accounting and bookkeeping software
• Expense management and reporting
• Mileage tracking and reimbursement
• Travel booking and management
• Receipt scanning and organization apps
• Data extraction and document understanding systems
• Business intelligence and data analytics
• Enterprise automation and RPA systems

Pricing

View detailed pricing

Support Options

• Email support
• Live chat support
• Phone support
• Ticketed support system
• Documentation and API reference
• Expert services for complex projects

Integrations

Cradl integrates with databases, cloud storage, RPA systems, and more:

• Postgres
• MySQL
• Microsoft SQL
• MongoDB
• Snowflake
• SAP
• Oracle
• AWS S3
• Microsoft Azure
• Google Cloud

API and SDKs

The models can be integrated via:

• REST API
• Python SDK
• JavaScript SDK
• Java SDK
• .NET SDK

Full API reference

What Customers Are Saying

"We're building the best software for accounting firms. Our customers have very high expectations, and with Cradl we can be confident that our document automation is best-in-class."

• Anders Skjetne, CEO of Centiga

"Cradl really convinced us by being so easy to use. We were quickly able to run projects independently, and the human-in-the-loop functionality enabled us to quickly deploy Cradl in our organization."

• Klaas Erik Römer, Posten

Frequently Asked Questions

How accurate are the models?

Accuracy ranges from 80-99% depending on your data. The more data you provide, the higher the accuracy.

How long does it take to train a model?

Training takes 5-10 minutes typically. The more data the longer it takes.

What file formats can I use?

JPEG, PNG images. PDF and TIFF support coming soon.

What languages are supported?

English, Norwegian, German, French, Spanish, Dutch and more.

Can I extract handwritten text?

Not currently. The models work best on computer-generated text.

Useful Links

• Sign up for free account
• Book a demo
• Documentation
• API reference
• Python SDK
• Pricing page