Title

AWS re:Invent 2023 - Use generative AI to name your pet after your favorite song (BOA208)

Summary

The session covered the use of generative AI to solve unique challenges, such as naming pets after favorite songs.
Brooke, Fred Gration, and Ben Cashman were the presenters, focusing on the ABCs of generative AI, building the solution, and prompt engineering, respectively.
They demonstrated how AWS services like Amazon Bedrock, CodeWhisperer, and Lambda can be used to build generative AI applications.
The session included live polling to engage the audience and gather input for generative AI demonstrations.
They showcased a proof of concept application that uses generative AI to suggest pet names based on song titles.
The application architecture involved Cognito for authentication, Lambda for serverless computing, DynamoDB for database storage, and CloudWatch for monitoring.
The presenters emphasized the importance of prompt engineering and provided insights into how to effectively communicate with generative AI models.
A blog post with detailed instructions on building the demonstrated application was announced.

Generative AI can be adapted to various business challenges beyond the playful example of naming pets.
The session highlighted the importance of understanding the basics of generative AI and how to interact with AI models through prompt engineering.
AWS offers a range of services that facilitate the development of generative AI applications, including managed services like Amazon Bedrock and developer tools like CodeWhisperer.
The presenters demonstrated the ease of integrating AWS services to create a generative AI application, emphasizing the use of APIs and IAM for secure and scalable solutions.
The concept of "temperature" in generative AI was discussed, which controls the creativity level of the AI's responses.
The session showcased the practical application of generative AI in a simple front-end application, highlighting the potential for more complex and production-ready use cases.
The presenters provided resources for attendees to learn more about generative AI and encouraged them to experiment with AWS services to build their own AI-driven solutions.