Rift ReKKap

Meet your analyst Vel'Koz
Matches you played
Your performance
Your signature position
Dragon Slayer
MVP match of the year

Inspiration

I was inspired by Genshin Impact’s yearly “Traveler’s Journey” recap, where players revisit their adventures through a personalized story. It made me realize that raw numbers don’t stick but stories do. I wanted to bring that same narrative magic to League of Legends, turning match data into something emotional and memorable. That idea became RIFT REKKAP.

What it does

RIFT REKKAP analyzes a player’s League of Legends match history and transforms it into a personalized, story-driven recap. Using AI and gameplay data, it highlights achievements, habits, and quirks through the voice of different champions, like Vel’Koz analyzing your “inefficient decision-making” or Teemo celebrating your chaos, well, for the future implementation. Some key features:

Fetches game data from the Riot API
Generates narrative summaries with AI-driven tone and insight using Amazon Bedrock
Visualizes trends such as win rate, favorite roles, and champion mastery
Outputs a shareable, end-of-season story that feels uniquely yours

How I built it

React, Node.js, AWS Bedrock with Amazon Nova Lite, League API, Kiro, Copilot

Challenges I ran into

Joining late: I found out about this hackathon pretty late, so I had to move fast, from brainstorming to building, with limited time to explore everything AWS Bedrock and League API offers.
League API limitations: At first, I didn’t fully understand what data was available in the League of Legends API. Fetching the exact match details I needed for deeper insights took a lot of trial and error.
AWS account issues: I ran into persistent permission and validation errors with my original AWS account. After spending hours debugging, I eventually created a new account just to get Bedrock working properly.
Finding meaningful insights: As a more casual gamer, I wasn’t sure which data points players who enjoy deep analysis would actually care about. I had to experiment with different metrics before finding ones that told interesting stories.

Accomplishments that I'm proud of

I’m proud that I built the entire pipeline from scratch, from Riot API data collection to AI-generated storytelling, all within like 3 days.

I managed to create AI summaries that genuinely feel like champion-narrated recaps, blending humor, data, and character voice. I also designed a clean, gamer-friendly interface that visually connects stats with storytelling, making data feel emotional rather than mechanical.

I'm pretty satisfied with what I've built.

What I learned

This project taught me how to:

Integrate AWS Bedrock models efficiently for real-time narrative generation.
Apply prompt engineering techniques to control tone and structure.
Manage API limits and optimize data retrieval pipelines for performance.
Use storytelling as a design tool to make analytical data feel meaningful. It also reminded me how creativity and technical experimentation can coexist. A recap doesn’t have to be numbers; it can be a memory.

What's next for RIFT REKKAP

Future updates will likely include:

More champion narrator personalities (like Teemo, and Heimerdinger).
More gameplay analytics, such as Skill Shots Hit, Skill Shots Dodged, and other fun, underrated metrics like number of times you bring down the first turrent, number of time you danced with the Rift Herald, that reflect playstyle beyond KDA.
Social sharing cards for Reddit, Discord, and X.
A mobile-friendly version so players can screenshot and share with their friends.
Fix the bugs:
- Occasionally, match data fails to load if the League API rate limit is hit too quickly.
- Total Damage Dealt To Champions doesn’t add up correctly.
- Time spent in the fountain displays inaccurately.
- “Most played friend” summoner information not showing correctly.

Ultimately, I want RIFT REKKAP to become a creative bridge between gameplay data and personal storytelling — where every match becomes part of a larger journey.

Why Vel'Koz?

I chose Vel’Koz as the first AI-agent to work on because he represents pure analysis and observation in the League of Legends universe. His lore revolves around studying life forms to understand them on a deeper, logical level. That made him the perfect “AI narrator” for RIFT REKKAP.