Buildsimple API Integration for Developers

Our cognitive Entity Extraction API allows you to extract business data from documents. Using Deep Learning, the Classification API can automatically detect the proper document type. If the results of entity extraction and classification are not satisfying these data can be validated and corrected with the Document Verifier.

Currently our APIs support German and English documents. With the free use of the Document Trainer, quality can be improved during extraction.

Entity Extraction API
Classification API
Document Verifier
Document Trainer

API Access & Documentations

Demo accesses and more information can be requested using our contact form

This is the Document Verifier

Document Verifier helps you validate and correct classified and extracted data. This way you can reduce your workload to a minimum.

Statement of work

blank
Entity Extraction API

1 Description of service

We offer Software as a Service solutions with modern cloud infrastructures. Our API services are designed to be integrated into the Cloud- and On-Premise solutions of our customers. ISR Information Products AG offers business- and technical services as Software as a Service subscriptions in the Amazon Web Services Marketplace. All of our services provides REST APIs for a simple integration. Our customers, partners and solution providers can use our services with low effort of investment due to cheap pay per use licenses.

Our products are based on the powerful Amazon Webservices, AWS for short, infrastructure and the associated scalability with hosting in the available regions of the AWS.

 

Buildsimple currently supports the following AWS regions:

Region Name Region Endpoint Protocol

EU (Frankfurt) eu-central-1 rds.eu-central-1.amazonaws.com HTTPS

USA East (Northern Virginia) us-east-1 rds.us-east-1.amazonaws.com HTTPS

 

The Entity Extraction API is a Business Service, to read pre-defined data from German and English documents. A complete list of all entities that are potentially extracted can be found in the API documentation. The service is using a deep learning based approach to identify the business type of each document to determine the correct neuronal net to read to business entities from the document. Supported business types of documents are invoices and contracts. The result of the extraction process is a JSON Stream that can be pulled from the AWS storage by a separate API call. Optionally we provide a callback functionality to our partners, to get notified immediately when the result was created, so that not polling for results is required.

 

1.1 Technology

We are using deep learning technologies to classify documents, to extract the entities and to validate the potential results. By this machine learning approach the quality of the extraction is not affected by different document layouts. Neuronal nets are based on weighted decision processes of trained models, so for our invoice and contract documents. Buildsimple was using a very large amount of different invoice and contract documents to achieve well trained models and a high extraction quality for new and unknown documents comparing to layout based extraction products. Ongoing training processes guarantee a continuous improvement of our extraction quality.

 

2 Service elements

2.1 Supported document types

  • Invoice

2.2 Supported file types

You can upload one of the following file types through the API: PDF, TIF, JPG. If a not supported files type was uploaded, the service is not able to create a valid result and the caller will receive an Error-Code. The file size cannot be more then 10 pages and 10 MB.

2.3 Classification

When a document is uploaded via the API, it will be automatically classified into one of the supported document types, listed in 2.1 or unknown. The Upload API allows the pre-specification of the document type by setting of a document type parameter. Details are described in the API documentation.

2.4 Languages

The Entity Extraction API supports the extraction of documents in German and English.

2.5 Quality of extraction

The extraction quality varies due to several factors

  • Quality of the images, scans or digital documents affects the OCR result which is the basis for the entity extraction
  • German and English documents results in different qualities
  • Because we do not know the input documents, we cannot guarantee a specific extraction quality

2.6 Result of extraction process

The result is provided as a JSON object for a duration of 8 hours and can be queried using a separate REST API call. After expiration of the 8-hour retention period, the extraction result is irrevocably deleted. Access to the extraction result is secured by authentication using AWS tokens and its connection to the customer’s or partner’s AWS account.

2.7 Storage of data

In addition to the storage of the extraction result for 8 hours mentioned in point 2.6, document information is stored temporarily during processing in order to carry out the various processing steps. The data stored temporarily for the processing of a document are irrevocably deleted immediately after the successful processing of a document. Should a technical error occur during the processing of a document, the temporarily stored data will be irrevocably deleted after a period of 1 hour. It is guaranteed that data will never be stored permanently after the processing of a document.

2.8 Encryption

The entire communication and processing chain within the AWS infrastructure is encrypted via SSL and HTTPS.

 

3 Conditions and obligations to cooporate

3.1 Amazon Web Services Account

The prerequisites for using the API is a valid AWS account and a prior registration of the account for the Entity Extraction API via the AWS Marketplace, as the usage billing is done via the AWS account.

3.2 Further prerequisites and obligations to cooporate

  • An active Internet connection is required to use the API. The customer / partner has access to the AWS cloud and its API gateway.
  • The customer / partner provides a competent and authorized contact person in case of support questions.
  • The customer / partner independently monitors the consumption of any concluded package contract. ISR Information Products AG does not proactively inform about reaching a package limit.
  • The customer / partner undertakes not to intentionally transmit defective documents to the API or to use the API in a deliberately defective manner.
  • The customer / partner is responsible for the transferred data / documents.

If one or more of the conditions stated here do not apply, ISR Information Products AG is not obliged to provide the described service with the agreed service levels.

These obligations to cooperate are generally provided in a quality that allows ISR Information Products AG to fulfill its contractual obligations without additional expenditure. ISR Information Products AG shall not be liable for delays in the provision of services and/or violations of the agreed service levels which are due to the non-performance of the cooperation obligations by the partner / customer or which are not attributable to ISR Information Products AG.

 

4 Terms of contract

The contract term varies depending on the billing model selected.

4.1 Pay as you go model

The pay-per-use billing model does not have a predefined contract term. The customer/partner pays exactly for each processed document. The contract can be ended at any time. A separate cancellation is not necessary.

 

5 Settlement

5.1 AWS Marketplace

Billing takes place directly via the AWS Marketplace, provided that the contract has been concluded via the AWS Marketplace. ISR Information Products AG transmits the number of processed documents hourly to the AWS Marketplace, so that accurate billing can be carried out.

5.2 Individual agreement

In the case of an individual contract agreement, invoicing is carried out via a monthly invoice from ISR Information Products AG. When using the pay-per-use model, ISR Information Products AG consolidates the documents processed in one month, which form the basis for the monthly invoice. In case of using a package contingent, the invoicing will be made directly on conclusion of the contract by invoice via the ordered package.

 

6 Licensing

Apart from licensing the Entity Extraction API, ISR Information Products AG does not charge any further license fees for the use of the API. The license is contained in the respective contract see point 5. All software products used are subject to the respective license conditions of the manufacturers.

 

7 Services

7.1 System Management & System Monitoring

Continuous monitoring of the system status by our DevOps team enables early detection of critical server system statuses. In this context, we carry out the following inspections, among others:

  • Continuous, central monitoring of services
  • Control of the system event log for critical system states and scaling of services
  • Control of hard disk utilization, processor and memory usage

We monitor the services using AWS CloudWatch. The services are scaled automatically by Amazon Web Services on the basis of parameters defined by ISR Information Products AG.

7.2 Customer & Partner Support

The license price includes access to the ISR Community, FAQs and a knowledge platform available to our customers and partners.

7.3 Call-acceptance & individual Support

https://www.isr.de/service/isr-product-support

7.3.1 Incident Management (Service interruption)

In the case of an incident, the service desk employee will perform the technical diagnosis and immediately try to find a solution with the help of a knowledge database. The successful solution as well as the solution steps are documented and the incident is closed after the fault has been rectified. The customer will be informed about the remedy of the fault. If an immediate solution is not possible, all previous measures are documented and the process is forwarded to subsequent instances (2nd level support or AMS). In the case of forwarding, the incident is proactively monitored by the service desk over the entire service period in order to ensure compliance with the agreed service levels on the basis of a defined escalation process. After successful resolution of an incident and closing of the ticket in the system, the customer will be informed.

 

8 SLA’s

8.1 Availability of the API

Despite the robust Amazon Web Services cloud infrastructure and a comprehensive quality assurance of the service by ISR Information Products AG, as well as uncontrollable quality gates, a 100% uninterrupted availability of the Entity Extraction API cannot be guaranteed. The software development and operating processes of ISR Information Products AG ensure a very high quality and stability of our services, so that we can reduce system downtimes to a minimum.

ISR Information Products AG cannot guarantee the availability of the API if all or part of the Amazon Web Services infrastructure is not available.

The Amazon Web Services Infrastructure enables ISR Information Products AG to install new versions, patches and bug fixes without interruption. An interruption of the service is not necessary.

 

9 Boundaries

ISR Information Products AG’s General Terms and Conditions of Business apply the current price list and service descriptions.

Classification API

1 Description of service

We offer Software as a Service solutions with modern cloud infrastructures. Our API services are designed to be integrated into the Cloud- and On-Premise solutions of our customers. ISR Information Products AG offers business- and technical services as Software as a Service subscriptions in the Amazon Web Services Marketplace. All of our services provides REST APIs for a simple integration. Our customers, partners and solution providers can use our services with low effort of investment due to cheap pay per use licenses.

 

Our products are based on the powerful Amazon Webservices, AWS for short, infrastructure and the associated scalability with hosting in the available regions of the AWS.

Buildsimple currently supports the following AWS regions:

Region Name Region Endpoint Protocol

EU (Frankfurt) eu-central-1 rds.eu-central-1.amazonaws.com HTTPS

USA East (Northern Virginia) us-east-1 rds.us-east-1.amazonaws.com HTTPS

 

The Classification API is a Business Service to classify your business documents into document types such as invoices and contracts. Our service offers an asynchronous API for document classification using two REST interfaces for document upload and result polling. The resulting document type is presented in JSON format. A list of supported document classes and languages can be found in the API documentation.

1.1 Technology

We are using deep learning technologies for document classification. Using this machine learning approach, the quality of the classification result is not affected by different document layouts. Neuronal nets are based on weighted decision processes of trained models, so for our invoice and contract documents. Buildsimple is using a very large amount of different invoice and contract documents to achieve well trained models and a high quality of the classification result for new and unknown documents compared to layout based classification products. Ongoing training processes guarantee a continuous improvement of our classification quality.

2 Service elements

2.1 Supported document types

  • Invoices

  • Contracts

2.2 Supported file types

You can upload single- and multi-page PDF files. If a not supported files type was uploaded, the service is not able to create a valid result and the caller will receive an Error-Code. The file size cannot be more then 10 pages and 5 MB.

2.3 Classification

When a document is uploaded via the API, it will be automatically classified into one of the supported document types, listed in 2.1 or unknown.

2.4 Languages

The Classification API supports the classification of documents in German and English.

2.5 Quality of classification

The classification quality varies due to several factors

  • Quality of the images, scans or digital documents affects the OCR result which is the basis for the document classification

  • German and English documents results in different qualities

  • Because we do not know the input documents, we cannot guarantee a specific quality of the classification result

2.6 Result of classification process

The classification result is provided as a JSON object for a duration of 72 hours and can be queried using a separate REST API call. After expiration of the 72-hour retention period, the classification result is irrevocably deleted. Access to the classification result is secured by authentication using AWS tokens and its connection to the customer’s or partner’s AWS account.

2.7 Storage of data

In addition to the storage of the classification result for 72 hours mentioned in point 2.6, document information is stored temporarily during processing in order to carry out the various processing steps. The data stored temporarily for the processing of a document are irrevocably deleted immediately after the successful processing of a document. Should a technical error occur during the processing of a document, the temporarily stored data will be irrevocably deleted after a period of 1 hour. It is guaranteed that data will never be stored permanently after the processing of a document.

2.8 Encryption

The entire communication and processing chain within the AWS infrastructure is encrypted via SSL and HTTPS. Document data that is stored temporarily is encrypted at rest using a combination of server- and client-side encryption.

3 Conditions and obligations to cooporate

3.1 Amazon Web Services Account

The prerequisites for using the API is a valid AWS account and a prior registration of the account for the Classification API via the AWS Marketplace, as the usage billing is done via the AWS account.

3.2 Further prerequisites and obligations to cooporate

  • An active Internet connection is required to use the API. The customer / partner has access to the AWS cloud and its API gateway.

  • The customer / partner provides a competent and authorized contact person in case of support questions.

  • The customer / partner independently monitors the consumption of any concluded package contract. ISR Information Products AG does not proactively inform about reaching a package limit.

  • The customer / partner undertakes not to intentionally transmit defective documents to the API or to use the API in a deliberately defective manner.

  • The customer / partner is responsible for the transferred data / documents.

If one or more of the conditions stated here do not apply, ISR Information Products AG is not obliged to provide the described service with the agreed service levels.

These obligations to cooperate are generally provided in a quality that allows ISR Information Products AG to fulfill its contractual obligations without additional expenditure. ISR Information Products AG shall not be liable for delays in the provision of services and/or violations of the agreed service levels which are due to the non-performance of the cooperation obligations by the partner / customer or which are not attributable to ISR Information Products AG.

4 Terms of contract

The contract term varies depending on the billing model selected.

4.1 Pay as you go model

The pay-per-use billing model does not have a predefined contract term. The customer/partner pays exactly for each processed document. The contract can be ended at any time. A separate cancellation is not necessary.

5 Settlement

5.1 AWS Marketplace

Billing takes place directly via the AWS Marketplace, provided that the contract has been concluded via the AWS Marketplace. ISR Information Products AG transmits the number of processed documents hourly to the AWS Marketplace, so that accurate billing can be carried out.

5.2 Individual agreement

In the case of an individual contract agreement, invoicing is carried out via a monthly invoice from ISR Information Products AG. When using the pay-per-use model, ISR Information Products AG consolidates the documents processed in one month, which form the basis for the monthly invoice. In case of using a package contingent, the invoicing will be made directly on conclusion of the contract by invoice via the ordered package.

6 Licensing

Apart from licensing the Classification API, ISR Information Products AG does not charge any further license fees for the use of the API. The license is contained in the respective contract see point 5. All software products used are subject to the respective license conditions of the manufacturers.

7 Services

7.1 System Management & System Monitoring

Continuous monitoring of the system status by our DevOps team enables early detection of critical server system statuses. In this context, we carry out the following inspections, among others:

  • Continuous, central monitoring of services

  • Control of the system event log for critical system states and scaling of services

  • Control of hard disk utilization, processor and memory usage

We monitor the services using AWS CloudWatch. The services are scaled automatically by Amazon Web Services on the basis of parameters defined by ISR Information Products AG.

7.2 Customer & Partner Support

The license price includes access to the ISR Community, FAQs and a knowledge platform available to our customers and partners.

7.3 Call-acceptance & individual Support

https://www.isr.de/service/isr-product-support

7.3.1 Incident Management (Service interruption)

In the case of an incident, the service desk employee will perform the technical diagnosis and immediately try to find a solution with the help of a knowledge database. The successful solution as well as the solution steps are documented and the incident is closed after the fault has been rectified. The customer will be informed about the remedy of the fault. If an immediate solution is not possible, all previous measures are documented and the process is forwarded to subsequent instances (2nd level support or AMS). In the case of forwarding, the incident is proactively monitored by the service desk over the entire service period in order to ensure compliance with the agreed service levels on the basis of a defined escalation process. After successful resolution of an incident and closing of the ticket in the system, the customer will be informed.

8 SLA’s

8.1 Availability of the API

Despite the robust Amazon Web Services cloud infrastructure and a comprehensive quality assurance of the service by ISR Information Products AG, as well as uncontrollable quality gates, a 100% uninterrupted availability of the Classification API cannot be guaranteed. The software development and operating processes of ISR Information Products AG ensure a very high quality and stability of our services, so that we can reduce system downtimes to a minimum.

ISR Information Products AG cannot guarantee the availability of the Classification API if all or part of the Amazon Web Services infrastructure is not available.

The Amazon Web Services Infrastructure enables ISR Information Products AG to install new versions, patches and bug fixes without interruption. An interruption of the service is not necessary.:

8.2 Service reporting

Regular service reports document the quality of our services. The reports contain details of the documents processed and their classification quality, taking into account the agreed SLAs, as well as details of trends or specific activities undertaken to improve service quality.

The service reports are made available to the customer / partner electronically at monthly intervals. The service report is standardized and has no customer-specific specifications.

It consists of the following contents:

  • Processed documents in the last settlement period

  • Availability of the service in the last billing period

9 Boundaries

ISR Information Products AG’s General Terms and Conditions of Business apply the current price list and service descriptions.

Document Verifier

Description of service

We offer Software as a Service solutions with modern cloud infrastructures. Our API services are designed to be integrated into the Cloud- and On-Premise solutions of our customers. ISR Information Products AG offers business- and technical services as Software as a Service subscriptions in the Amazon Web Services Marketplace. All of our services provides REST APIs for a simple integration. Our customers, partners and solution providers can use our services with low effort of investment due to cheap pay per use licenses.

 

Our products are based on the powerful Amazon Webservices, AWS for short, infrastructure and the associated scalability with hosting in the available regions of the AWS.

Buildsimple currently supports the following AWS regions:

Region Name Region Endpoint Protocol

EU (Frankfurt) eu-central-1 rds.eu-central-1.amazonaws.com HTTPS

USA East (Northern Virginia) us-east-1 rds.us-east-1.amazonaws.com HTTPS

 

Document Verifier extracts all relevant data from unstructured documents. A dialog enables the verification and correction of the extracted data for quality assurance purposes. After approval by the auditor, the system makes the specialist data available to subsequent systems such as ERP, CRM or specialist applications via an API.

The integration into your processes will advance your company in the digital transformation and secure a decisive competitive advantage.

 

Service elements

Document receipt

The Document Verifier accepts image documents via a simple REST API, e.g. from a scanner, e-mail box or specialized application. Alternatively, a user, e.g. the examiner himself, can upload individual documents.

Supported document types

  • Invoice
  • Contract

Supported file types

  • For files with more than 10 pages, only the first 10 pages are analyzed.
  • The files may be up to 5 MB in size.
  • The following formats are supported: PDF, TIF, JPG.

 

Verification dialog

The documents are classified and technical properties such as addresses, names, amounts are extracted. Which data is recognized depends on the classification. The classification determines the document type with the domain-oriented properties.

The recognized data is clearly displayed in a dialog next to the original document. During the check, the selected properties are highlighted in the document so that the checker can compare them and correct the missing or incorrect data manually. Each verified document is re-introduced into the Entity Extraction API training data, continuously improving extraction quality.

Provision of verified data

After approval by the auditor, the system provides the document with the properties via a REST API to the subsequent systems such as ERP, merchandise management, accounting and CRM systems as well as specialist applications.

User administration and inbox

User management allows an administrator to register and manage additional users for the system. A group inbox helps to distribute the documents to be verified.

User profile

Every registered user has a user profile that contains all personal settings as well as the contractual and legal documents. Users can manage their accounts and delete their accounts.

Languages

The Document Verifier can be used in German and English.

Data storage

After a document has been checked by a user, the document is deleted. The result of the verification process remains to be used for training the Entity Extraction API. This data contains the OCR text and the corrected data units of the verified documents.

Encryption and data security

The entire communication and processing chain within the AWS infrastructure is encrypted using SSL and HTTPS.

 

Conditions and obligations for cooperation

Amazon Web Services Account & Entity Extraction API Registration

The prerequisite for using Document Verifier is a valid AWS account and prior registration of the Entity Extraction API account through the AWS Marketplace. Access information to the Document Verifier is provided to the user during the Entity Extraction API registration process.

Further prerequisites and obligations for cooperation

An active Internet connection is required to use Document Verifier. The customer has a valid account for the Document Verifier.
For support questions, the customer provides a competent and authorized contact person.
The customer independently monitors the consumption of a concluded parcel contract. ISR Information Products AG does not proactively inform about reaching a package limit.
The Customer undertakes not to intentionally transmit any defective Documents to the API or to intentionally misuse the API.
The customer is responsible for the transferred data / documents.
If one or more of the conditions stated here should not apply, ISR Information Products AG is not obliged to provide the service described with the agreed service levels.

As a rule, these cooperation obligations are provided in a quality that enables ISR Information Products AG to fulfil its contractual obligations without additional effort. ISR Information Products AG shall not be liable for delays in the provision of services and/or breaches of the agreed service levels which are attributable to the non-fulfilment of the cooperation obligations by the customer or which are not attributable to ISR Information Products AG.

 

Terms & Conditions & Licensing

Document Verifier is free to use and available exclusively to all registered Entity Extraction API customers. When a customer logs off from the Entity Extraction API, access to the Document Verifier also becomes invalid and all user data is deleted.

Document Trainer

1 Description of service

We offer Software as a Service solutions with modern cloud infrastructures. Our API services are designed to be integrated into the Cloud- and On-Premise solutions of our customers. ISR Information Products AG offers business- and technical services as Software as a Service subscriptions in the Amazon Web Services Marketplace. All of our services provides REST APIs for a simple integration. Our customers, partners and solution providers can use our services with low effort of investment due to cheap pay per use licenses.

 

Our products are based on the powerful Amazon Webservices, AWS for short, infrastructure and the associated scalability with hosting in the available regions of the AWS.

Buildsimple currently supports the following AWS regions:

Region Name Region Endpoint Protocol

EU (Frankfurt) eu-central-1 rds.eu-central-1.amazonaws.com HTTPS

USA East (Northern Virginia) us-east-1 rds.us-east-1.amazonaws.com HTTPS

 

The Buildsimple Document Trainer is an optional user interface to train the document extraction quality of the Entity Extraction API.

 

2 Service elements

2.1 Upload documents

You can upload one of the following file types through the API: PDF, TIF, JPG. If a not supported files type was uploaded, the service is not able to create a valid result and the caller will receive an Error-Code. The file size cannot be more then 10 pages and 10 MB.

2.2 Browse uploaded documents

When documents are uploaded, a user can browse the uploaded documents in an inbox to select one document to be trained by the user.

2.3 Train documents

Documents can be trained by reviewing the extracted data of the Entity Extraction API and correct the missing or wrong extracted data manually by the user. The corrected data is submitted to Buildsimple to be used in the next training process for the Entity Extraction API engine.

2.3.1 Supported document types

Invoice

2.4 User profile

Each registered user has a user profile which contains all personal settings, as well as the contractual and legal documents. User are able to manage their accounts, as well as deleting their accounts.

2.5 Languages

The Buildsimple Document Trainer can be used in German and English.

2.6 Storage of data

After a document was trained by a user, the document will be deleted. The result of the training process is persisted to be used für the training of the Entity Extraction API. This data contains of the OCR text and the corrected data entities of the trained documents.

2.7 Encryption

The entire communication and processing chain within the AWS infrastructure is encrypted via SSL and HTTPS.

 

3 Conditions and obligations to cooporate

3.1 Amazon Web Services Account & Entity Extraction API registration

The prerequisites for using the Buildsimple Document Trainer is a valid AWS account and a prior registration of the account for the Entity Extraction API via the AWS Marketplace. The access information to the Buildsimple Document Trainer is submitted to the user during the registration process of the Entity Extraction API.

3.2 Further prerequisites and obligations to cooporate

An active Internet connection is required to use the Buildsimple Document Trainer. The customer has a valid account for the Buildsimple Document Trainer.
The customer provides a competent and authorized contact person in case of support questions.
The customer independently monitors the consumption of any concluded package contract. ISR Information Products AG does not proactively inform about reaching a package limit.
The customer undertakes not to intentionally transmit defective documents to the API or to use the API in a deliberately defective manner.
The customer is responsible for the transferred data / documents.
If one or more of the conditions stated here do not apply, ISR Information Products AG is not obliged to provide the described service with the agreed service levels.

These obligations to cooperate are generally provided in a quality that allows ISR Information Products AG to fulfill its contractual obligations without additional expenditure. ISR Information Products AG shall not be liable for delays in the provision of services and/or violations of the agreed service levels which are due to the non-performance of the cooperation obligations by the customer or which are not attributable to ISR Information Products AG.

4 Terms of contract & Licensing

The usage of the Buildsimple Document Trainer is free of charge and exclusively available to all registered Buildsimple Entity Extraction API customers. If a customer unsubscribes from the Entity Extraction API, the access to the Buildsimple Document Trainer will be invalidated as well and all user data will be deleted.