Skip to main content

Integration Architecture with FPT AI eKYC System via SDK

1. Introduction​

This document describes the system architecture of two deployment options when customers integrate with the FPT AI eKYC system using Web or Mobile SDK, including architecture diagrams, data flow details, advantages and disadvantages, and important considerations during implementation.

  • FPT AI eKYC Backend: The backend system of FPT AI eKYC, responsible for managing eKYC sessions and data flows in and out of the system.
  • FPT AI eKYC SDK: SDK provided by FPT AI eKYC for web and mobile environments, allowing integration with customer websites or mobile applications.
  • Proxy Server: An intermediary server between the customer's system and the FPT AI eKYC system, located in the customer's infrastructure, responsible for data transmission between both parties.
  • Customer Backend: The customer's backend system, responsible for receiving data generated during eKYC processes or other logic related to eKYC flows.

3. Integration Options​

3.1. Integration without Proxy Server​

3.1.1. Description​

This is the simplest integration option, where the FPT AI eKYC SDK sends data directly to the FPT AI eKYC backend without going through a Proxy Server. To ensure data security and safety, processing results will be sent to the customer's backend through a pre-registered callback URL.

The data flow diagram between components in this option is as below:

In addition to obtaining eKYC session processing results from callback requests, customers can get these results from the SDK returned to the application or actively retrieve them from the FPT AI eKYC backend via the GET /callback/get_result API using the client_uuid (details are described in the API_result documentation).

3.1.2. Advantages​

  • Simple, easy to deploy. The FPT AI eKYC system and customer's system are almost completely separate.
  • Reduces costs in terms of time and effort for the customer's technical team during implementation.
  • No need to maintain an additional intermediary server. Therefore:
    • Minimizes latency in data transmission between SDK and FPT AI eKYC backend.
    • Easier to handle/detect errors that arise during integration and eKYC system usage.

3.1.3. Disadvantages​

  • Cannot control the data flow between SDK and FPT AI eKYC backend.
  • Cannot intervene in the final results of the eKYC session, completely dependent on FPT AI eKYC system processing results.
  • Retrieval of eKYC processing results and data depends on FPT AI eKYC system storage on FPT Cloud.

3.2. Integration using Proxy Server​

3.2.1. Description​

This is a more complex integration option, where the FPT AI eKYC SDK sends data to the customer's Proxy Server, which then forwards the data to the FPT AI eKYC backend. Processing results will be sent back to the Proxy Server, which then forwards them to both the SDK and the customer's backend for storage.

The data flow diagram between components in this option is as follows:

eKYC session data and results can be processed and stored from the Proxy Server (within the customer's system) without waiting for callback from the eKYC Backend. Customers can still retrieve callback results if needed. Additionally, eKYC data and results can also be obtained from the SDK returned to the application.

3.2.2. Advantages​

  • Can control the data flow between SDK and FPT AI eKYC backend.
  • Can intervene in the final results of the eKYC session without being completely dependent on FPT AI eKYC system processing results.
  • eKYC session data and results pass through the customer's system first, allowing complete control over storage without depending on FPT AI eKYC system cloud storage.

3.2.3. Disadvantages​

  • More complex to implement. The FPT AI eKYC system and customer's system are not completely separate, and eKYC flow stability heavily depends on the connection to the Proxy Server.
  • Requires maintaining an additional intermediary server. Therefore:
    • Increases latency in data transmission between SDK and FPT AI eKYC backend.
    • More difficult to handle/detect errors that arise during integration and eKYC system usage.
  • Requires an additional technical team to maintain the intermediary server.