Question 1

What is Cambrion?

Accepted Answer

Cambrion is an agentic AI platform for automating document workflows. It uses AI agents to transform unstructured documents, such as invoices and reports, into structured data for use in your business systems.

Question 2

What company is behind Cambrion?

Accepted Answer

The platform is provided by Cambrion GmbH, a technology company founded in Munich, Germany, in 2024. The company focuses on agentic AI for data structuring in Europe. The industry analyst firm Deep Analysis recommends buyers in document-heavy industries such as logistics, healthcare, and insurance to consider Cambrion for its innovative approach to document processing.

Question 3

What is intelligent document processing (IDP)?

Accepted Answer

Intelligent Document Processing (IDP) automates the extraction of information from documents and loads it into business systems. Cambrion uses an agentic AI approach, which differs from earlier IDP systems and OCR technology, which often relied on fixed templates.

Question 4

How does Cambrion's AI work?

Accepted Answer

The platform is based on Vision-Language Models (VLMs) and an agentic AI core. The process to set up a workflow and automate processing follows four distinct steps: Goal Definition: A user provides a sample document and specifies the required data fields. This tells the AI what information to look for. Pipeline Creation: Cambrion’s Pipeline Agent analyzes the request once and creates a unique processing pipeline. The proposed data schema can then be configured. Pipelines can also include additional steps, such as validating data against internal data sources or enriching it via web search. API Integration: The configured pipeline is integrated into your existing systems via a dedicated REST API endpoint. This connects the automated workflow to your software environment. Continuous Processing: Once integrated, the pipeline runs automatically. New documents are sent to the API, where Cambrion processes them and delivers the extracted, structured data back to your systems in a machine-readable format.

Question 5

Does Cambrion replace OCR?

Accepted Answer

Cambrion's technology differs from traditional Optical Character Recognition (OCR). While OCR is designed to convert text in an image into digital text, Cambrion's VLM-based approach is designed to interpret the document's structure and the context of the data. This allows it to flexibly process varied and complex layouts without pre-configured templates.

Question 6

Which use cases does Cambrion support?

Accepted Answer

The platform is designed to be horizontal and supports two primary types of use cases: automating internal business processes and enhancing software products. Automating Internal Processes: Companies use Cambrion to automate document-centric workflows across various departments. Common applications include the processing of commercial invoices, delivery documents, supplier reports, patent filings, and HR records. Product Integration: Software providers and SaaS companies can embed Cambrion's capabilities directly into their own applications. This allows them to offer powerful, native document AI features to their customers without having to build the core technology. Cambrion’s technology is capable of addressing a wide range of document automation challenges. If you have a unique use case, please contact us.

Question 7

Which document formats can be processed

Accepted Answer

Cambrion can process PDF, Word, Excel, PowerPoint, and common image files (JPEG, PNG, etc.). The system is capable of handling documents with a high page count. We continuously enhance our document format capabilities. Contact us if you have a specific file format that you intend to process.

Question 8

Can Excel files be processed?

Accepted Answer

Yes. The Cambrion platform can process Excel files, including support for multiple sheets within a single file.

Question 9

Can Cambrion handle diverse layouts?

Accepted Answer

Yes, the Cambrion platform is specifically designed to handle dynamic document layouts and formats, eliminating the need for static template configuration. This capability is central to its functionality, ensuring adaptability to ever-changing documents.

Question 10

Is Cambrion able to process handwriting?

Accepted Answer

Yes, the technology can recognize and process both machine-generated text and handwriting.

Question 11

What languages does Cambrion support?

Accepted Answer

Cambrion generally supports all languages that use a variation of the Latin script. Our customers most frequently process documents in German, French, Italian, English, Danish, Norwegian, Swedish, Finnish, Spanish, Portuguese, Czech, Slovak, Polish, Hungarian, and Romanian. Additionally, Cambrion can process other character sets, such as Chinese, Hindi, Japanese, and Korean. The platform's user interface is currently available in English.

Question 12

How many documents are needed to start?

Accepted Answer

A single sample document, along with your natural language extraction requirements, is enough to configure a workflow. Large datasets and manual data labeling are not required for setup.

Question 13

How long does it take for Cambrion to process a document?

Accepted Answer

Our REST API supports both real-time (synchronous) and batch (asynchronous) processing to fit your workflow needs. The average processing time can vary based on several factors, including the number of pages, file size, the complexity of the document's layout, and the specific functions configured in your pipeline, such as data validation or enrichment steps.

Question 14

How does Cambrion integrate with existing software?

Accepted Answer

Cambrion provides several integration options: REST API: The primary method of integration is a direct connection to the platform's REST API. Low-Code Platforms: Cambrion connects with various low-code platforms. A certified connector is available for Make, and customers have also successfully integrated the platform using OutSystems. Partner Integrations: Our network of partners can provide and support integrations with specific systems. We are continuously expanding our library of connectors to ensure straightforward compatibility with common ERP, WMS, CRM, TMS, and HR systems.

Question 15

How long does it take to implement Cambrion?

Accepted Answer

You can implement Cambrion in less than 1 day. With the help of Cambrion’s Pipeline Agent, your custom pipeline is set up in minutes. The technical API integration typically takes just a few hours. A complete enterprise onboarding process usually takes between one and four weeks, with project timelines often dependent on the client's internal processes.

Question 16

Do I need technical expertise to use Cambrion?

Accepted Answer

No. The platform includes a user interface that allows non-technical users to configure and manage data extraction pipelines. A full API is available for developers requiring more control and configuration. Integrations such as with our Technology Partner Make allow for a seamless no-code integration into existing workflows.

Question 17

Which output formats are possible?

Accepted Answer

Data is provided in JSON format via the API. The platform also supports Markdown. For testing purposes, users can manually export results as Excel files.

Question 18

Where can I find the REST API documentation?

Accepted Answer

The API documentation is available at http://docs.cambrion.io/

Question 19

Is my data secure with Cambrion

Accepted Answer

Yes. Cambrion is ISO 27001 certified and adheres to the General Data Protection Regulation (GDPR). The platform's architecture was designed to comply with EU data privacy and security standards. Various deployment options ensure that usage complies with your own information security requirements.

Question 20

Where is my data processed and stored

Accepted Answer

All customer data is processed and stored on servers located within the European Union. Our strategic cloud infrastructure partner is T Cloud Public (formerly Open Telekom Cloud) by T-Systems in Germany.

Question 21

Do you train AI models with my data?

Accepted Answer

No. Customer data is not used to train any AI models, neither Cambrion’s nor the general-purpose models from our underlying technology partners. Your data is processed exclusively to fulfill your document processing request.

Question 22

Can Cambrion be used in my private cloud or on-premise?

Accepted Answer

Yes, container-based private cloud deployment and on-premise and options are available for organizations with specific data governance policies.

Question 23

How much does Cambrion cost?

Accepted Answer

Cambrion is offered via a flexible subscription model. The subscription fee includes platform access for unlimited users, a defined volume of pages for processing, and all ongoing platform updates. Final pricing is based on factors such as the deployment mode, your total annual document volume, pipeline functionality, and the required level of support. Please schedule a call to discuss your specific requirements and receive a detailed quote.

Question 24

What is your onboarding process like?

Accepted Answer

The onboarding process includes a use case review, a guided UI walkthrough, and API setup support to ensure a smooth start.

Question 25

Can I try the platform before committing?

Accepted Answer

Yes, we offer free trials and can arrange a proof-of-value pilot to evaluate the platform with your specific documents and use case.

Dokumente rein.
Belastbare Daten raus.

Datenextraktion scheitert oft dort, wo Dokumente komplex werden.

Ein einziger Agent für jedes Layout.

In drei Schritten zur Automatisierung.

Ein Beispiel bereitstellen.

Pipeline ist vorkonfiguriert.

Ihre Dokumente verarbeiten.

Wie es funktioniert.

Ein Beispiel bereitstellen.

Pipeline ist vorkonfiguriert.

Ihre Dokumente verarbeiten.

Produkterprobt und praxiserprobt.

3× schnellere Rechnungsverarbeitung

1.800 Stunden jährliche Ersparnis

Von 30 Minuten auf eine reduziert

10.000 Patentseiten pro Jahr

Wesentliche Vorteile.

Bereitstellungsoptionen.

Sichere EU-Cloud

Private Cloud

On-Premise

Überzeugen Sie sich selbst anhand Ihrer eigenen Dokumente.

Dokumente rein. Belastbare Daten raus.

Datenextraktion scheitert oft dort, wo Dokumente komplex werden.

Ein einziger Agent für jedes Layout.

In drei Schritten zur Automatisierung.

Ein Beispiel bereitstellen.

Pipeline ist vorkonfiguriert.

Ihre Dokumente verarbeiten.

Wie es funktioniert.

Ein Beispiel bereitstellen.

Pipeline ist vorkonfiguriert.

Ihre Dokumente verarbeiten.

Produkterprobt und praxiserprobt.

3× schnellere Rechnungsverarbeitung

1.800 Stunden jährliche Ersparnis

Von 30 Minuten auf eine reduziert

10.000 Patentseiten pro Jahr

Wesentliche Vorteile.

Bereitstellungsoptionen.

Sichere EU-Cloud

Private Cloud

On-Premise

Überzeugen Sie sich selbst anhand Ihrer eigenen Dokumente.

Dokumente rein.
Belastbare Daten raus.