Case study – Manufacturing
A generative-AI storytelling product, in production on AWS.
Little Legends turns a child into the hero of their own storybook: parents upload a photo and answer a few questions, and AI generates a personalised book. 56k.Cloud designed and built the platform on AWS – scalable, secure, and shaped for rapid experimentation with different AI models.
The problem
Little Legends needed to launch a personalised-storytelling service quickly and meet market demand – on infrastructure that could scale to high volumes of image and text generation, and flex as they experimented with different AI models to refine the experience.
- rapid development and deployment to reach market fast
- scale to high volumes of image and text generation
- freedom to swap and experiment with AI models
- security and data privacy – built against AWS's own quality framework
The architecture
A serverless core – API Gateway, Cognito, and Lambda – scales with demand and keeps user data protected. A custom Stable Diffusion XL model on Amazon SageMaker generates images tailored to each child; Claude 3.5 Sonnet on Amazon Bedrock produces the narratives and image prompts. The whole solution was designed and reviewed against the AWS Well-Architected Framework.
The outcome
It runs in production today – built with the Little Legends team, owned by them, and operated by us for as long as they want.
- personalised storybook service launched in 12 weeks
- serverless infrastructure that scales with demand, no servers to manage
- rapid experimentation across AI models to keep refining the experience
- security and data privacy built to the AWS Well-Architected Framework
It runs in production today – built on AWS and shaped for rapid experimentation with new AI models.