semantic player search impl

[DIodide]

Find Talent: AI-Powered Player Discovery for Coaches

Summary

Add a new "Find Talent" page to the coach dashboard (/dashboard/coaches/find-talent) that enables coaches to search for players using natural language queries. The system will use vector embeddings for semantic search and Google Gemini for generating player analysis content.

Architecture

flowchart TB
    subgraph frontend [Frontend - Coach Dashboard]
        SearchBar[Natural Language Search]
        Filters[Quick Filters]
        ResultsGrid[Player Results Grid]
        Modal[Player Detail Modal]
    end

    subgraph api [tRPC API Layer]
        SearchEndpoint[talentSearch.search]
        AnalysisEndpoint[talentSearch.getAnalysis]
    end

    subgraph services [Backend Services]
        EmbeddingService[Embedding Service]
        GeminiService[Gemini Service]
    end

    subgraph database [PostgreSQL + pgvector]
        Players[(players table)]
        Embeddings[(player_embeddings table)]
    end

    subgraph external [External APIs]
        GeminiAPI[Google Gemini API]
    end

    SearchBar --> SearchEndpoint
    Filters --> SearchEndpoint
    SearchEndpoint --> EmbeddingService
    EmbeddingService --> GeminiAPI
    EmbeddingService --> Embeddings
    SearchEndpoint --> Players
    ResultsGrid --> Modal
    Modal --> AnalysisEndpoint
    AnalysisEndpoint --> GeminiService
    GeminiService --> GeminiAPI

Loading

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Review	Updated (UTC)
eval-next	Error		Jan 17, 2026 9:41am