Serverless project using Boto3 and Python Lambda functions to automate interaction with Amazon Rekognition to analyze videos. Rekognition label detection analysis is trigger by uploading a video file to configured S3 bucket. The results of the analysis are stored in a DynamoDB table. Video files need to be in mp4 format.
Install serverless framework using npm
npm install serverless -g
Install pipenv to manage a virtualenv and dependencies
pip3 install pipenv
Create config.dev.json file based on config template and specify aws profile to use, desired S3 bucket name and DynamoDB table name. Serverless framework will create those resources during deployment.
To deploy run
serverless deploy