Tactician's Fundamentals

Built as a capstone for UCLA Anderson's Science and Strategy of AI (MGMT298D), Tactician's Fundamentals is a full-stack coaching tool that compares your TFT games against a dataset of 140,000+ Challenger and Grandmaster matches. The goal: make the decision-making patterns of high-elo players legible to everyone else.

Most TFT tools give you stats. This one tries to explain what to do with them. After learning about retrieval-augmented generation and how models can be grounded in domain-specific context, I wanted to build an AI that could reason about TFT the way a high-elo player does — rather than repeat generic tier-list advice.

What it does

• Stat Calculator — unit stats across star levels and item combinations, with item recommendations derived from Challenger win-rate data
• TFT Coach — enter your Riot ID to compare your economy and leveling decisions against high-elo benchmarks
• Gods — full Set 17 god boons and trials reference
• Ask AI — open-ended Q&A with live search grounding and a structured TFT knowledge base

How it works

The reasoning layer uses a RAG-style approach: each response is grounded at inference time in champion data, trait breakpoints, item stats, and pro-level benchmarks rather than training data alone. Three models work in concert — Gemini 2.5 Flash for coaching and live search, Groq LLaMA 3.3 for champion tips and fallback responses, and Riot's Match API for personal game history.