RiftSaga

Inspiration

The main inspiration behind RiftSaga comes from the iconic narrative scenes of the Lord of the Rings prologue — a powerful blend of epic storytelling, emotional depth, and cinematic atmosphere. We wanted to capture that same sense of grandeur and mythic scale, transforming each player's League of Legends journey into a tale worthy of the legends of Runeterra. By merging data-driven insights with narrative artistry, RiftSaga turns gameplay statistics into a living story — evoking the feeling of being part of an epic saga that unfolds uniquely for every summoner.

What it does

RiftSaga is an interactive storytelling experience that transforms a player’s League of Legends journey into a personalized, cinematic adventure. Through a 14-step narrative enhanced with dynamic visuals, players progress alongside an AI companion system that reacts in real time, offering champion-specific commentary and personalized responses.

The platform also features AI coaching, delivering tailored strategy insights and performance recommendations based on gameplay data. Beyond storytelling, content generation tools allow users to create custom summoner posters and cinematic highlight reels that capture their in-game achievements.

Finally, with integrated social sharing, players can easily showcase their personalized stories and visuals across platforms — celebrating their League year with friends and the wider community.

How we built it

RiftSaga is powered by a combination of robust backend and frontend technologies. The backend, built with FastAPI, handles League of Legends analytics and conversational AI agents via the Strands SDK, with AWS Bedrock providing the inference engine. The frontend, developed with React + Next.js and hosted on CloudFront, delivers a seamless, interactive experience.

Key features include an AI-powered poster generator for custom visual summaries, machine learning playstyle analysis using KMeans clustering and PCA to categorize players into archetypes like "carry" or "support", and intelligent agents capable of analyzing gameplay patterns and providing personalized coaching. The platform automatically fetches match history from Riot’s API, processes thousands of data points to generate tailored improvement recommendations, and stores everything in AWS S3 for scalability. Social relationship analysis and all content are accessible through a production-ready API with proper error handling and CORS support.

Challenges we ran into

During development, we encountered several hurdles:

• AWS account restrictions temporarily blocked access to critical services.
  
• IAM permissions issues complicated the creation of the AWS Bedrock knowledge base.
  
• Token limits were exceeded when building AI agents with the Strands SDK, requiring careful optimization.

Accomplishments and lessons learned

We’re proud to have built a fully functional, end-to-end application leveraging multiple AWS services, including S3, Bedrock, Strands AI Agent SDK, Bedrock Knowledge Base, CloudFront, and ACM. Through this project, we gained deep experience in cloud architecture, AI integration, and scalable application design, as well as hands-on knowledge of deploying production-grade AI systems.

What’s next for RiftSaga

Looking ahead, our focus will be on improving performance and scalability:

• Optimizing Strands agents for faster, more efficient inference.
  
• Developing an Infrastructure-as-Code pipeline to automate knowledge base creation and management.

Built With

• amazon-web-services
• bedrock
• cloudfront
• fastapi
• python
• react
• s3
• terraform