Today, companies utilize a wide variety of serverless services offered by major cloud providers as well as emerging platforms like Vercel (a cloud specialized in hosting web front-end applications) and CloudFlare (a global network with numerous security and hosting services). These services can be grouped into three main categories:
FaaS: a cloud service that relieves developers from managing their own infrastructure and is an event-driven execution model (e.g., AWS Lambda, Azure Functions, Google Cloud Functions, Cloudflare Workers)
Containers: Software packages that contain all the elements needed for execution in any environment (e.g., AWS ECS Fargate, Azure Container Apps, Google Cloud Run, Fly.io)
Web apps, es: Cloudflare Pages, Netlify, Vercel
Simultaneously, specialized development frameworks and infrastructure-as-code are also gaining traction, making this model highly effective in terms of productivity. The main ones include Serverless Framework (multi-provider), Architect (AWS), and Ampt (AWS).
A very interesting report on the state of serverless is the one periodically produced by DataDog: https://www.datadoghq.com/state-of-serverless/
The usage of managed services for running serverless containers is experiencing significant growth due to the launch of new services. Leading in this area is Google Cloud, which with the launch of Cloud Run as early as 2019, made containers its flagship serverless product, already widely adopted today.
This type of service simplifies adoption and migration towards serverless since companies often already use containerized environments for their applications. Container-based services also support a wider range of languages and have less stringent limitations compared to FaaS services.
From what we see daily, there are many use cases suitable for serverless adoption in the TechFin domain, and many projects would benefit from it in terms of maintainability and costs.
The most common scenario currently for hosting new applications in this sector is to use container orchestrators on proprietary infrastructure, with RedHat Openshift being the predominant choice in most cases. Among the architectural patterns, microservices usage prevails, with standard Web development technologies oriented towards implementing JSON APIs and Single Page Applications (JavaScript). These two elements integrate very well with current serverless-supporting technologies.
It is a fact that the migration to the Cloud is underway, and in the near future, the doors will be open for the adoption of Serverless services. What seems most likely is that we will remain within the perimeter of the major Cloud providers (AWS, GCP, and Azure) and that those serverless services that allow integration on private VPC networks will be used.
Looking further into the future, edge services could cover critical and widespread use cases in the TechFin domain. One particular example is applications supporting Forex trading.