Based in Argentina, InvoiTrade, is the first electronic SaaS platform for the Negotiation of Electronic Credit Invoices, which operates according to article 13 of the Productive Financing Law (27,440). This platform is aimed mainly at all Argentine SMBs, with the mission of bringing them the best financing options, allowing the financial inclusion of this sector.
InvoiTrade is a company of the InvoiNet group, a founding member of the Argentine Chamber of Fintech, and a leading company in the management of electronic invoices, throughout Latin America, processing more than 1 million invoices per month, with more than 100 corporate clients and more than 200 thousand users of connected companies.
Its team is made up of professionals from the technology and financial sector, who provide an excellent service to all its clients. Its clients are among the most important national and international companies.
They have the support of the World Bank through the International Finance Corporation (IFC), which is one of its shareholders and the main international public institution, exclusively for the private sector in developing countries.
Amazon S3, AWS DynamoDB, Amazon Rekognition, Amazon API Gateway, AWS Lambda, AWS SAM, AWS CloudFormation.
In order to preserve the security of its platform, InvoiTrade decided to incorporate a biometric authentication solution into its product. The said solution had to be integrated with its current user authentication and authorization module, which has been in operation since the beginning of the project.
Incorporating the biometric authentication workflow meant reconciling the new image data, hosted in an Amazon Rekognition collection and on Amazon S3, with the original user database, hosted by a third-party cloud provider. In turn, the creation of APIs that could be used by the application to carry out the validity process of the users within the flow and communicate the status to the original backend (also hosted in the same cloud provider).
First, considering that their entire backend was hosted by a third-party provider, we had to organize sessions to identify the IDs of the data to carry out the integration with the biometric identification process in AWS. For this, we reconciled that the “PersonID” was going to be the main data by which we would associate the user of the original backend with our biometric flow.
On the other hand, we decided which APIs we were going to develop so that both the mobile application and the website could incorporate this flow into their On-Boarding and Login processes.
Finally, we decided to develop the AWS Lambda functions and deploy them in an automated way using the AWS Serverless Application Model, that way we could deploy function code and definition of APIs in an automated and consistent way.
The APIs we decided to create were:
POST /index-faces – Responsible for receiving a photo, detecting faces, verifying that a single face exists, and saving it to AWS DynamoDB and Amazon Rekognition.
POST /search-faces – Responsible for receiving a photo and searching the Amazon Rekognition and AWS DynamoDB collection for matches.
DinoCloud decided to use AWS SAM as the main deployment and versioning framework for the solution. Meanwhile, we unify the deployment of the entire product (application, configuration, and infrastructure) in a unique and simple way.
Regarding the development of functions, we use the Python language for the backend of the solution.
A next-generation serverless architecture.
With the help of DinoCloud, InvoiTrade managed to integrate the 100% serverless biometrics solution into its authentication and authorization workflow, where we decided to allocate the core of the solution to Amazon Rekognition, a reliable solution from the AWS suite of services.
InvoiTrade incorporated the flow in its mobile application and in its web platform, from the DinoCloud side we provide support in the use of APIs to integrate it into the user experience of the product client.
In the production deployments, the flow of On-Boarding and Biometric Login is already incorporated together with the proof of life of the users. The serverless solution allows the client to scale over time without the need to incorporate changes in the current architecture.