Serverless Experiments
mlops
project
tutorial
The Application
Update (October 2026): both applications have since been retired, so the live demos are no longer available. The methodology write-ups linked below remain.
I recently built an application using serverless architecture, which is described in more detail here. There were two running experiments:
- Pun Generator Application, methodology described here.
- Entity Resolution Application, methodology described here.
What is Serverless?
AWS explains the term serverless, aka Function as a Service (FaaS) as:
A serverless architecture is a way to build and run applications and services without having to manage infrastructure... You no longer have to provision, scale, and maintain servers to run your applications, databases, and storage systems.
The figure below summarizes a generic serverless architecture with AWS. S3 and Dynamo are stateful components for storage that may or may not be needed in all application. Lambda is the serverless, event-driven compute. Instead of spinning up a machine that is latent most of the time, this is on demand compute. Hence, for a few pennies a month (actually free!), I can have the live application running. Amplify is the front end for these applications.