Jerry Hargrove

Jerry Hargrove is a cloud architect, developer and evangelist who guides companies on their journey to the cloud, helping them to build smart, secure and scalable applications for their businesses. Jerry has an extensive background in software architecture, development and operations and brings with him over 20 years of experience as a developer, architect & manager for companies like Rackspace, AWS, Intel and now, Lucidchart.

Workshop: How to Build a Serverless ML Enabled Bot

During this tutorial, we’ll build a serverless Slack bot to help identify & classify images using Machine Learning services on AWS. This will include creating and configuring an AWS Lambda-backed Slack bot for interacting with users, and configuring & using Amazon Rekognition and SageMaker for detection and classification.


At the conclusion of this tutorial session, you will have gained familiarity with AWS Machine Learning services and gained experience using AWS SageMaker for image classification. You will leave the tutorial with a working bot implementation that you can continue to use, modify, and refine if desired. This experience and knowledge can be used to facilitate your own Machine Learning projects.


Some experience with AWS is required. In particular, familiarity with setting up IAM users, configuring and testing Lambda functions, using CloudWatch logs, and configuring other AWS services and features would be very helpful.


We will using Python to implement the Lambda function and will be walking through Python code examples, so some familiarity with the language would be beneficial.


Requirements for the workshop

This session will require an active AWS account and that you have IAM permissions to administer AWS services and features, in particular, Lambda, IAM, SageMaker, and Rekognition. If you do not have an AWS account, you should signup for one before attending this session here:


Important: We will be using SageMaker ml.p3.2xlarge instances for model training during the workshop. Prior to attending this workshop, you should submit a request to AWS support to increase your limit of ml.p3.2xlarge SageMaker training instances to a minimum of 2 for the Frankfurt region (or another region you choose to use). 


Note: The AWS services used during this session will incur some costs to your AWS account. AWS credit vouchers will be given to those who attend this tutorial to compensate for these costs.


In addition to the AWS account, you will also need a Slack workspace with permissions to create and administer a Slack app. If you do not have a Slack Workspace, you should sign up for one before attending the session here:


/ Platinum




/ Climate

/ Diversity

/ Hosted by