This project is a full demo featuring Anomaly Detection models on cookies from developing the notebooks to the web app deployment.
MAIN FEATURES
- Three models (Baseline, EfficientAD, FOMO AD)
- A web app
- Efficient AD model using Sagemaker serverless endpoint
- FOMO AD model using your browser as real-time inference from Edge Impulse
- An API for the serverless inference endpoint
- Automatic deployment with AWS using Terraform
.github
code related to the deployment of the website after a git pushai
Code related to the model's developmentapi
Code the inference APIinfrastructure
Code related the automated infrastructurewebsite
Code related to the web app
-
Deploy the website after a git push IF the commit message contains the text: "deploy:website" within the commit message
- branch dev to deploy to dev.anomaly.parf.ai
- branch prod to deploy to anomaly.parf.ai