Because I’m API – Discover the 5 pillars of our technology

In a rapidly expanding FinTech like Xpollens, the services offered and the technology upon which they rely are evolving at lightning speed. The challenge lies in reconciling agility, stability, and high availability. During an interview, Mathieu Taberlet, CTO of Xpollens, guides us through the foundations of our technical stack and the strategic choices that underpin the success of our platform.

At Xpollens, technology is primarily the work of a team of talents. Their mission? To design, develop, and maintain a state-of-the-art Banking-as-a-Service platform, adhering to the best development practices as well as organisational standards, and prioritising system security from the outset, which is crucial in our payment domain.

The technical team is organised into squads – architecture, security, engineering, monitoring – and employees are assigned to functional domains. This structure allows them to gain domain expertise and become experts. This organisation is essential for facilitating collaboration within the team, as well as with business teams and clients. That’s why we pay particular attention to the profile of our talents. The desire to learn and curiosity are essential for continuously improving their knowledge, but also, and most importantly, for mastering a domain end-to-end and being confident in their design.

The CI/CD model, standing for Continuous Integration and Continuous Delivery/Deployment, enables us to streamline and accelerate the development cycle of platform services. The automation and structure brought by this model optimise our entire chain, from development templates to deployment phases, including validation checks, security protocols, and cost management.

This practice also involves integrating code changes much more regularly and in smaller increments. Two immediate benefits emerge:

• Reduced time between the completion of development and its deployment into production.
• Decreased impact in case of issues. Deployment becomes less stressful for both technical and business teams.

The efficiency of our CI/CD approach is a key element of the maturity of our technical stack and the performance of our platform.

At Xpollens, the technical stack, comprising the set of technologies enabling our processes, is predominantly Microsoft-based, but not exclusively so…

API development relies on the .NET Core framework for enhanced performance during project compilation and execution.
We prioritise being consistently on the latest Long-Term Support versions of the framework to benefit from its advancements in security and optimisation, ensuring we remain free from any obsolescence effects.

Hosting is done on the Azure public cloud, with a preference for Serverless architecture.
The priority is to ensure our applications run smoothly while allowing our developers to focus on code rather than server management. Leveraging Azure expertise and Serverless architecture frees us from server management burdens, providing:

• Native integration,
• Automatic optimisation of consumed resources,
• Improved scalability of the platform,
• Cost control.

We primarily use Linux for its cost-effectiveness in terms of licenses and its lightweight operating system that consumes fewer resources. Additionally, Linux offers enhanced security, performance, and compatibility.

Thanks to Infrastructure as Code (IaC), we manage our infrastructure resources just like we manage our application code, using a common language, the same controls, and structure.
In alignment with our stack, choosing the Bicep language ensures us:

Greater responsiveness: we are closely aligned with the Azure platform with zero delay between a Microsoft update and its implementation.

Increased reliability: we reduce the risks of errors or overwritten code with smoother and more efficient script optimisation and application.

IaC and Bicep enable us to:

• Create resources more quickly,
• Ensure compliance with our provisioning and architecture policies for end-to-end control,
• Empower developers through templates,
• Remain compatible with other clouds such as AWS and Google Cloud Platform.

Another choice that significantly contributes to the overall performance of our technology, NoSQL databases, being more permissive and flexible, allow us to remain agile and continue working at any scale. Capable of handling large volumes of rapidly evolving unstructured data, they are particularly effective at managing unpredictable data, often with extremely fast query speeds. They are indispensable in highly innovative environments like ours and those of our clients.

At the core of our offering, APIs enable our clients’ applications to access the functionalities and data of other services or systems. In simple terms, they allow for easy integration of white-label services into their own environment.
The benefits of using APIs include:

Increased interoperability: the development of our APIs adheres to the entire set of REST architectural constraints and relies on SWAGGER documentation standards, making communication between heterogeneous systems possible, regardless of the technologies or programming languages used. This promotes collaboration between different stakeholders, facilitates data exchange and integration by our clients.

Enhanced security: APIs allow for controlling access to data and functionalities by providing authentication and authorisation mechanisms. This helps protect resources and ensures data confidentiality.

In addition to APIs, we combine a microservices architecture and webhooks for asynchronous event management.
This set of components contributes to making our platform easier to scale and maintain.
However, the microservices architecture imposes two crucial steps:

• Upstream, the client must understand its own workflows as a prerequisite for defining the events to be run.
• Downstream, monitoring must oversee the cascade of events.

It’s no longer just monitoring a single application, but monitoring hundreds of microservices and the communication between these microservices. That’s why we also rely on the expertise of a Site Reliability Engineer. He enhances operations once the code is deployed in production, focusing on operations and maintaining highly available services.

When the ambition is to become the European leader in Banking-as-a-Service, we equip ourselves with a modern and high-performing IT infrastructure, summarised in 5 points:

• A robust CI/CD pipeline for rapid deployment of developments.
• A solid framework that updates automatically.
• Reliable and flexible hosting.
• Tailored monitoring for microservices.
• And last but not least, the best for the end, a team of happy talents 😉.