# TRUE Amazon Personalize Workshop July 26th, 2021 # Lab Accounts #### AWS Event Engine Login: https://dashboard.eventengine.run/login #### Event Hash: 9d55-13f2c8e604-34 ##### Github Page: [AWS Retail Demo Store](https://github.com/aws-samples/retail-demo-store) # Lab Guide ## Prerequisite 1. Login in to Event Engine accounts with URL and event hash above. 2. Search for the SageMaker service in the AWS console. 3. Navigate to Notebooks and click the jupyterlab link 4. Navigate into workshop folder and the personalize folder 5. Launch the personalize 1.1 jupyter notebook ## Lab Schedule: ## 1. Lab Overview - Jupyter (20 minutes) Nick * Lab schedule overview * Intro to jupyter notebooks * Why running all? - this notebook takes 1 hour to complete * We are going to train a model its gonna take time * AWS resource creation: CLI, SDK, or GUI * Creating User/Item Interactions * After data is created plots are generated * This is all fake data that is created * Graphs help to understand the data * Leahi hotel demo * show the retail demo website (no product recommendations) > Note: cloudformation URL - cloudfront is AWS CDN (no HTTPS**) * have users create a login and select a persona ## 2. Personalize Datasets (10 min) Stason * Describe Dataset group and hierarchy for datasets * we define a schema and then import data * personalize requires one datasets - interactions * users and items metadata are optional but can help your model learn ## 3. Personalize Solutions (15 min) Stason * AWS console * Show personalized dashboard in console * Show datasets and schema * 3 types of Recipe - tell the Amazon.com story > * 1. personalized & popularity recipe > * 2. personalized reranked items based on user and item > * 3. similar Items * Solution includes a recipe and some other pieces of information to train an ML model behind the scenes ## 4. Personalize Campaign (10 min) Nick * show console 3 campaigns are creating (correspond to the 3 solutions/recipes) * show the web application in the console: * show demo guide in retail site * ec2 elastic loadbalancers * ecs cluster, ecs services, ecs tasks * this demo is using microservices running on ECS - some are python and some are Golang * show test campaign results in console with a user id ## 5. Retail Demo Website (10 min) Stason * have users go back to the retail demo website * guide them through the retail demo store * Show Chat, Search, SMS - not functioning due to other labs. * Checkout Item in DEMO site # Survey #### [Survey Link](https://survey.immersionday.com/rmtuspWnR) We take feedback very serious and it is one of our mechanisms for constantly improving. We strongly encourage you to complete this brief 1 minute [survey](https://survey.immersionday.com/rmtuspWnR). # Appendix #### [AWS Certified Machine Learning Specialty](https://aws.amazon.com/certification/certified-machine-learning-specialty/) #### [AWS Machine Learning University](https://aws.amazon.com/machine-learning/mlu/) #### [AWS Machine Learning Embark Program](https://aws.amazon.com/ml-embark/)