Discovery
Our discovery process is intended to help organisations that are investigating Serverless architecture to understand where they can achieve specific benefits through adopting Serverless architectures. This process will help make the business case, and provide a starting point on where to start a Serverless project, whether greenfield or a migration.
There are many case studies, vendor white papers, blog posts and conference talks on the benefit of Serverless, but that needs to be contextualised into how that could benefit your organisation. The discovery processes output is a report that provides specific detail on where your organisation will benefit, and what applications/services are most suitable to be built at Serverless applications.
The process consists of 3 parts
Scoping to understand how broad the discovery process should be
Information gathering to collate all the required data, both qualitative and quantitative
Synthesis and reporting on serverless readiness
The process is intended to be lightweight and iterative, with a focus on providing insight as quickly as possible. It is time boxed to a maximum of one month, and scoped appropriately. Due to the 1 month time box, there could potentially be multiple discovery cycles, with each cycle diving deeper to achieve more granularity, and each cycle providing actionable data.
This approach is preferred over potential big bang multi-month exhaustive discovery exercises, as we believe Serverless adoption should be iterative and not a big bang approach.
Scoping
Before we start the process we need to define the scope quite tightly to prevent it becoming an open ended engagement that is all encompassing. For us our discovery process runs for a maximum of one month, with defined outcomes and scope agreed during this phase. Elements that can impact the scope include
How well and detailed your environment is documented
Availability of monitoring, tracing and log data
How many applications your want assessed
Complexity of your applications and environments
Scoping will typically be a video call with some information shared before hand.
Information Gathering
Once the scope is agreed, we’ll move into information gathering. We take a qualitative and quantitative approach to information gathering.
On the qualitative side we will interview relevant individuals, typically developers, architects, DevOps engineers, product managers, and engineering managers to understand what needs to be addressed.
The quantitative side will involve analysing observability data, including but not limited to metrics, traces and logs, to understand how existing applications work. We will also analyse your code base, architecture and documentation where possible.
Synthesis and Reporting
Once we have enough information to satisfy the agreed scope we’ll review that information and formulate a number of hypothesis on where adopting serverless will provide benefits. We’ll either test these hypotheses, or provide supporting evidence to show how you will benefit.
An example of hypothesis testing could include creating a small demo application to send traffic to, recreating the typical request pattern your applications may have, to show a serverless application will behave with similar traffic.
Where it’s not possible to test a hypothesis easily, we’ll provide supporting evidence from external sources, with detail on how that relates to your specific case.
Finally we will collate this into a single report you can share widely, up to senior or executive management and with junior team members, to help build the case for serverless in your organisation, and provide a starting point.