Serverless deployment of ML on AWS and Anomaly Detection

  Machine learning

Serverless Deployment of ML on AWS - Andrej Hoos and Šimon Soták (Represent)
Serverless computing is a new and exciting paradigm for running software in the cloud. In recent years, it has gained a lot of traction thanks to the popularity of services like AWS Lambda and Google Cloud Functions. Serverless solves many headaches that come with traditional hosting on virtual servers like automatic scaling, OS updates, and health monitoring.
Deploying a machine learning model to a serverless platform like AWS Lambda comes with some obstacles, though. We will talk abou the limitations of AWS Lambda for ML with Python and how we overcame them. We are releasing a small open-source package for compiling your Python ML dependencies for deployment to AWS Lambda

Intro to Anomaly Detection - Martin Barus (
The abstract will be added

Language: English

Andrej Hoos and Šimon Soták (Represent)
Andrej Hoos and Šimon Soták are Machine Learning Engineers at Represent, one of the most successful Czechoslovak startups to date. Their focus is on automation in the printing world using computer vision and machine learning.

Martin Barus (

- 17:45 - 18:00 - Your arrival
- 18:00 - 18:40 - Serverless Deployment of ML on AWS
- 18:40 - 18:50 - Short break
- 18:50 - 19:30 - Intro to Anomaly Detection
- 19:30 - 22:00 - Networking in Bitcoin Coffee