This project uses Streamlit to build example generative AI applications on AWS with Amazon Bedrock focused on digital marketing use cases.
- Product Metadata Generator: Use product images to generate product descriptions, feature lists, and meta tag code for Open Graph, schema.org, and more...
- Product Review Summarizer: List the most common positive and negative comments from a set of consumer product reviews and summarize the overall sentiment.
- Product Blog Writer: Generate a 750-1000 word blog post with a comma-separated list of related SEO keywords.
- Doument FAQ Generator: Use PDF documents to generate a list of expected customer questions and their answers.
- Product Assistant Chatbot: Have a text and image-based conversation with a Claude 3 chatbot.
Follow these steps to install and run this project on AWS using the AWS Cloud9 cloud-based integrated development environment (IDE).
-
Create a new Cloud9 environment.
- Name:
gen-ai-demo
- EC2 Instance:
m5.large
(recommended) - Platform:
Ubuntu Server 22.04 LTS
- Name:
-
Once created, open the new Cloud9 environment.
-
Toggle off AWS managed temporary credential in Cloud9 by going to Preferences > AWS Settings > Credentials
-
Configure the AWS CLI with your permanent AWS credentials. You will need an
AWS Access Key ID
andAWS Secret Access Key
for the next step.- (Optional) Create an access key for an IAM user with AdministratorAccess if you don't already have one.
-
Configure your AWS credentials and region in the Cloud9 terminal using the AWS CLI.
aws configure
- Use the integrated Terminal to clone this GitHub repository.
git clone https://github.com/robsable/amazon-bedrock-genai-streamlit
cd amazon-bedrock-genai-streamlit
- Install Python requirements.
pip3 install -r setup/requirements.txt -U
- Run the Streamlit app.
cd app
streamlit run Main_Menu.py
- In Cloud9, go to the Preview menu and select Preview Running Application. A new tab in the Cloud9 IDE will open and load your running application.
Once the app is up and running, you can begin to customize for your own use cases.
-
Edit main entry page content for the app in
app/Main_Menu.py
. -
Add your own Streamlit scripts to the
app/pages
directory. -
Edit main menu items and layout in
app/.streamlit/pages.toml
.
- Delete the Cloud9 environment you created.