Skip to main content
Back to Projects
In developmentFlagship project

StratForge AI

Full-stack Counter-Strike match tracking, economy modeling, and strategy recommendations

Next.js 15React 19TypeScriptPrismaPostgreSQLSupabaseClerkVercel
View repository Open live application
Next.js 15
Frontend
React 19
UI Runtime
PostgreSQL
Data Layer
Vercel
Deployment

Project overview

StratForge AI models rounds, economy, roles, utility, and strategy records in a Next.js application. Implemented work includes live match tracking, economy prediction, recommendations, authentication, persistence, and APIs. Search, journaling, administration, and match-data ingestion remain roadmap work.

The problem

Competitive Counter-Strike strategy depends on economy, side, roles, utility, history, and current round context, but that knowledge is fragmented across videos, notes, and team memory.

The approach

Modeled match context and strategy structure in a full-stack application so it can track a live match, reason about the situation, and return explainable recommendations from persistent data.

Implemented systems

  • Live Match Engine and Round Tracker for maintaining current match context
  • Economy Prediction and Strategy Recommendation engines grounded in round and team state
  • AI Coach and match dashboard for situational recommendations
  • Database-backed strategies with steps, utility requirements, default roles, tags, and statistics
  • Clerk authentication with Prisma models, PostgreSQL/Supabase persistence, and API endpoints
  • Zustand client state and TanStack Query server-state workflows

System flow

A simplified view of the major system boundaries
  1. Next.js UI
    Match view
  2. Match State
    Rounds
  3. Coach Engine
    Recommend
  4. API + Prisma
    Services
  5. PostgreSQL
    Strategies

Architecture

01.Interface: Next.js 15, React 19, Tailwind CSS, and shadcn/ui
02.Client state: Zustand for live match state and TanStack Query for server data
03.Logic: round tracking, economy prediction, recommendation engine, and AI coach
04.API and auth: Next.js endpoints protected through Clerk
05.Persistence: Prisma models backed by PostgreSQL and Supabase
06.Delivery: Vercel deployment for the full-stack application

Engineering challenges

1

Representing live round context without losing history needed for later recommendations

2

Structuring economy, role, utility, and strategy data for explainable coaching

3

Separating implemented workflows from future ingestion and analysis concepts

Current state

  • Live match, round tracking, economy, recommendation, coach, dashboard, strategy, auth, and persistence foundations are implemented
  • Strategy records support steps, utility requirements, default roles, tags, and performance statistics
  • The application is deployed on Vercel and continues to evolve

Roadmap - not yet implemented

Utility Library and Economy CalculatorMatch Journal, favorites, and global searchAdmin PanelDemo and match-data ingestionHLTV and broader competitive-match analysis concepts

Continue exploring

Azure Retail Data Pipeline

A 300K-order Azure pipeline from Data Factory and Azure SQL to Power BI

Starter Toolkit

PowerShell diagnostics and snapshot tooling built from recurring enterprise IT support work