# MatchMind

> MatchMind is a free AI football intelligence web app for World Cup 2026 fans — powered by Gemini, MongoDB Atlas, and Google Cloud Agent Builder.

MatchMind helps World Cup 2026 fans get analyst-grade football intelligence from natural-language questions. It is a real agent — not a chatbot wrapper — where MongoDB is the intelligence layer and Gemini is the reasoning layer on top.

## Product Summary

- **Name:** MatchMind
- **URL:** https://app.matchmind.xyz/
- **Category:** AI football intelligence / sports analytics
- **Event focus:** FIFA World Cup 2026
- **Tagline:** Know Your Game. Own Every Moment.

## Core Capabilities

1. **Stats** — Top scorers, group standings, player and team statistics
2. **Predictions** — Match outcome analysis with confidence signals
3. **Fantasy** — Lineup and player selection insights
4. **Tactical** — Team strengths, weaknesses, and matchup analysis
5. **Historical** — Head-to-head records and tournament context

## Example Questions

- Who are the top scorers in Group B?
- Predict the Brazil vs France quarterfinal
- Build me a fantasy XI from Group C players
- What's Argentina's defensive weakness this tournament?
- What is the head-to-head record between Brazil and France?

## Architecture (Short)

```
Fan question → Next.js /agent → /api/agent
  → Gemini classifies intent (stats | prediction | fantasy | tactical | historical)
  → query_football_data MCP tool → MongoDB Atlas
  → Gemini analyzes records → structured JSON analyst card
```

## Tech Stack

| Layer | Technology |
|-------|------------|
| Framework | Next.js 16+ App Router |
| Agent brain | Gemini (gemini-2.5-flash-lite) |
| Orchestration | Google ADK (@google/adk) |
| Database | MongoDB Atlas |
| MCP tools | lib/mcp.ts |
| Deploy | Vercel |

## Public Pages

- [Home](https://app.matchmind.xyz/): Landing page and product overview
- [Football Intelligence Agent](https://agent.matchmind.xyz/): Interactive Q&A agent
- [Architecture](https://app.matchmind.xyz/docs/architecture): Technical documentation

## What We Do Not Do

- Official broadcast or real-time feed replacement (independent MongoDB intelligence platform)
- Betting or gambling recommendations
- Guaranteed match predictions

## Frequently Asked Questions

### What is MatchMind?

MatchMind is an AI football intelligence agent for World Cup 2026 fans. It classifies natural-language questions, queries a MongoDB Atlas football database via MCP tools, and returns structured analyst-grade responses with key stats, confidence signals, and follow-up suggestions.

### What can I ask MatchMind?

You can ask about World Cup 2026 stats, match predictions, fantasy lineups, tactical analysis, and head-to-head history. Example questions include top scorers by group, quarterfinal predictions, fantasy XI builds, and team weaknesses.

### How does MatchMind work?

MatchMind uses a four-step pipeline: Gemini classifies your question intent, an MCP tool queries MongoDB Atlas with aggregation pipelines, Gemini reasons over the retrieved records, and the UI renders a structured analyst card with headline, analysis, stats, and a follow-up prompt.

### What technology powers MatchMind?

MatchMind is built with Next.js, Gemini (gemini-2.5-flash-lite), MongoDB Atlas, Google Cloud Agent Builder (ADK), and TypeScript. It deploys on Vercel and uses an MCP tool interface in lib/mcp.ts for database retrieval.

### Does MatchMind use live FIFA broadcast data?

MatchMind is powered by live MongoDB sync after tournament kickoff — independent of any official broadcast data feed. Before World Cup kickoff (11 June 2026), it shows clearly labeled preview mockup data for demo UX. After kickoff, real results sync via npm run sync or the admin agent and appear as live MongoDB data.

### Who built MatchMind?

MatchMind was built by Mojeeb Titilayo (@mojeebeth) at BlindspotLab for the Google Cloud Rapid Agent Hackathon 2026, MongoDB Partner Track.

### Is MatchMind free to use?

Yes. The fan-facing MatchMind agent at agent.matchmind.xyz is free to use. MatchMind is open source on GitHub at github.com/mojeebdev/matchmind.

## Contact

- Builder: [Mojeeb Titilayo](https://x.com/mojeebeth) (@mojeebeth)
- Organization: [BlindspotLab](https://blindspotlab.xyz)
- Source: [GitHub](https://github.com/mojeebdev/matchmind)

## AI Discovery Files

- [LLM summary](https://app.matchmind.xyz/llms.txt)
- [Expanded LLM context](https://app.matchmind.xyz/llms-full.txt)
- [FAQ for answer engines](https://app.matchmind.xyz/faq-ai.txt)
- [AI interaction policy](https://app.matchmind.xyz/ai.txt)
- [Structured identity](https://app.matchmind.xyz/identity.json)
- [Sitemap](https://app.matchmind.xyz/sitemap.xml)
- [Robots rules](https://app.matchmind.xyz/robots.txt)