Building Serverless Machine Learning Models in the Cloud
summary
Builds and deploys actual Python ML models to AWS Lambda using scikit-learn, wires them behind API Gateway, and stores serialized models on S3 so Lambda pulls updated versions. Presents a real sentiment binary classifier, a production pipeline with periodic offline retraining, and a multi‑model deployment (≈50+ models at Cloud Academy) that works in production but needs hacks for native libraries, suffers cold starts, and hits the 5‑minute execution and packaging limits.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:16User-centered ML needs cloud
- 4:45Model deployment ownership
- 7:55Traditional deployment challenges
- 12:10Serverless architecture for ML
- 15:46Serverless benefits for ML
- 19:22Cloud Academy serverless ML
- 24:06Serverless data pipeline
- 27:41Serverless ML limitations
- 32:58Serverless tools ecosystem
- 34:04Q&A
related talks
01h 27mDeploying a Machine Learning Model to the Cloud Using AWS Lambda
44mPragmatic Machine Learning in the Cloud
12mServerless Deep Learning
34mDeploying a Machine Learning System That Only Runs Locally
45mServerless Machine Learning: a beginner's guide
34mBuilding scalable end-to-end deep learning pipelines in the cloud
01h 21mBuilding a Machine Learning Enabled Bot in the Cloud
32mAutomating Production Production Level Machine Learning on AWS