Weekly Digest 2
Café IO's weekly digest to update you with awesome stuff written and published this week. This week, we discuss 'Models as Serverless functions' and 'Data Cascades in high stakes domain'.
Links to the articles published this week:
We hope you had a great week!
We are here with a quick refresher of the posts this week on Cafe IO.
Subscribe to “Coffee and Engineering” to stay updated with fresh content.
Everyone wants to do the Model work not Data Work
We reviewed a paper “Everyone wants to do the Model work, not Data Work” published by Google earlier this year. The authors highlight that high-stake domain like health and wildlife conservation demand high-quality data. The current emphasis on model accuracy neglecting the domain expertise and data quality are causing a data cascade.
Data cascades have been defined by the authors as:
“Compounding events causing negative, downstream effects from data issues, that result in technical debt over time”.
Recent trends in AI have indicated a large emphasis on modelling and increasing accuracy while neglecting data engineering. Alongside, the domain expertise in AI is not considered that important, as modelling and accuracy are.
The authors In the high-stakes domain, it becomes imperative to lay down rules on data quality and on establishing proper feedback loops for the entire AI data lifecycle.
Read the complete paper summary here.
ML Models as Serverless Functions
Machine learning Systems has a constant feedback loop of receiving new data continuously which is needed for a retraining cycle. This is an additional complexity which adds to the list of operational challenges. Serverless patterns enable a decoupled design and promote microservice-based architecture enabling separation of concerns which helps towards maintainability. At the same time, they are not the silver bullet and carry the drawbacks of microservice architecture.
Serverless architectures are application designs that incorporate third-party “Backend as a Service” (BaaS) services, and/or that include custom code run in managed, ephemeral containers on a “Functions as a Service” (FaaS) platform. Serverless functions enable coders to deploy their machine learning models without worrying a lot about maintaining servers and choosing between options.
To help you get started with Serverless functions, we published an article containing an introduction to the model as serverless functions, compared managed vs hosted solutions and PaaS (Platform as a service) vs Serverless and looked at GCP and AWS Lamba as useful options.
Check out the complete article here.
List of articles published last week at Cafe IO:
2 part series to help you get started and advanced with git: Part 1 and Part 2
Coffee break article on approaching ML problem
We will bring some awesome stuff next week also. At Cafe IO, we believe in delivering better than yesterday and we would like you to become a part of our journey. Please subscribe and share the blog with your friends so that they can get to know about us.
Do Like, Share, Subscribe and Invite your Friends! Have a happy weekend.