README.md

metadata

title: 404final
emoji: ⚙️
colorFrom: blue
colorTo: green
sdk: gradio
app_file: app.py
pinned: false

EGR404Project

Crafting VALORANT Tactics with AI.

Description

This will be an AI-powered tool that generates custom team based strategies in the free-to-play video game known as VALORANT that are both map and agent specific and can adapt to desired playstyle. The community that surrounds VALORANT is both lively and extremely competitive, yet there are little to no tools that provide such functionality other than generic guides from the internet (usually YouTube). While it is a niche idea, it is something highly valuable to this community that requires little knowledge and/or effort to use.

Resources

Riot Games API (Riot Developer Portal) for player stats and agent data, LLMs for strategy generation, YouTube breakdowns of professional strategies for my own comparison, and custom user feedback at the individual and team level to evaluate the success of AI generated strategies.

Deliverables

At the end of the project, I plan to show demonstrations of this tool in real time as well as my gathered user experiences. A rough timeline will be: 1. Gathering successful strategies from my resources (YouTube, Reddit, VALOPLANT [https://valoplant.gg/]) 2. Making a prototype 3. Testing with agents from known successful strategies to test for similarities, differences, innovation, and ability to adjust and/or adapt 4. Reaching out to users to test for functionality, correctness 5. Adjusting the prototype as needed during these findings 6. Reaching out to users again to adapt the provided strategies 7. Evaluating their successes and/or struggles, shortcomings, etc. 8. Refine and repeat until sufficient 9. Final presentation