> For the complete documentation index, see [llms.txt](https://senselab.gitbook.io/senselab-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://senselab.gitbook.io/senselab-docs/docs/diagrams/03-ciclo-vida-dato-ia.md).

# Ciclo de Vida del Dato con IA — 10 Servicios

> Cómo un archivo físico pasa por OCR, se clasifica automáticamente con CAByS y alimenta el Análisis Financiero y Credit Scoring.

## Diagrama de Flujo Completo

```mermaid
graph LR
    subgraph INPUT["📥 ENTRADA"]
        PDF["📄 Factura<br/>PDF / IMG"]
        TXT["📝 Descripción<br/>Producto"]
        HIST["📊 Datos<br/>Históricos"]
        MSG["💬 Consulta<br/>Usuario"]
    end

    subgraph STAGE1["🔍 ETAPA 1 — Extracción"]
        OCR["🤖 OCRService<br/>───────────<br/>Gemini Vision<br/>92% Precisión<br/>───────────<br/>Extrae:<br/>• Proveedor<br/>• Líneas detalle<br/>• Totales<br/>• Confianza 0-100"]
    end

    subgraph STAGE2["🏷️ ETAPA 2 — Clasificación"]
        CABYS["🏷️ CabysClassifier<br/>───────────<br/>Gemini + CABYS DB<br/>98% Precisión<br/>───────────<br/>12 categorías<br/>10,000+ subcódigos<br/>Código tributario CR"]
    end

    subgraph STAGE3["🧠 ETAPA 3 — Análisis"]
        PRED["🔮 PredictionService<br/>Forecast demanda 30d<br/>Alertas reabastecimiento"]
        ANOM["🚨 AnomalyDetection<br/>Detección fraude 95%<br/>Transacciones mayor a 3σ"]
        CREDIT["💳 CreditScoring<br/>Score 0-100<br/>6 factores ponderados"]
    end

    subgraph STAGE4["💡 ETAPA 4 — Generación"]
        CHAT["💬 ChatbotService<br/>RAG + Intent Detection<br/>Asistente contextual"]
        CONTENT["📝 ContentGenerator<br/>Emails automáticos<br/>Reportes narrativos"]
    end

    subgraph STAGE5["📊 ETAPA 5 — Salida"]
        FE["📋 Factura Electrónica<br/>Datos pre-llenados"]
        INV["📦 Inventario<br/>Alertas stock"]
        DASH["📈 Dashboard KPIs<br/>Métricas tiempo real"]
        REPORT["📄 Reportes<br/>PDF / Excel / CSV"]
        ALERT["🔔 Notificaciones<br/>Webhooks + Email"]
    end

    subgraph PROVIDERS["☁️ Proveedores IA"]
        GEM["🟡 Google Gemini 2.0<br/>GRATUITO — 15 RPM"]
        GPT["🟢 OpenAI GPT-4o<br/>Fallback de pago"]
    end

    PDF --> OCR
    TXT --> CABYS
    HIST --> PRED
    HIST --> ANOM
    HIST --> CREDIT
    MSG --> CHAT

    OCR --> CABYS
    CABYS --> FE
    CABYS --> INV

    PRED --> DASH
    PRED --> INV
    ANOM --> ALERT
    ANOM --> DASH
    CREDIT --> REPORT
    CREDIT --> DASH

    CHAT --> REPORT
    CONTENT --> ALERT

    OCR -.-> GEM
    CABYS -.-> GEM
    CHAT -.-> GEM
    CONTENT -.-> GEM
    PRED -.-> GEM
    ANOM -.-> GEM
    CHAT -.-> GPT

    style STAGE1 fill:#1a472a,stroke:#2ecc71,color:#fff
    style STAGE2 fill:#1a3a5c,stroke:#3498db,color:#fff
    style STAGE3 fill:#4a1a5c,stroke:#9b59b6,color:#fff
    style STAGE4 fill:#5c3a1a,stroke:#e67e22,color:#fff
    style STAGE5 fill:#1a1a5c,stroke:#e74c3c,color:#fff
```

## 10 Servicios de IA — Detalle

| #  | Servicio                    | Provider                | Endpoints             | Precisión | Descripción                        |
| -- | --------------------------- | ----------------------- | --------------------- | --------- | ---------------------------------- |
| 1  | **GeminiService**           | Google Gemini 2.0 Flash | LLM genérico          | —         | Motor principal, **GRATUITO**      |
| 2  | **OpenAIService**           | GPT-4o                  | Fallback + embeddings | —         | Alternativa de pago                |
| 3  | **OCRService**              | Gemini Vision           | 3 endpoints           | **92%**   | Escaneo inteligente de facturas    |
| 4  | **ChatbotService**          | Gemini 2.0 Flash        | 3 endpoints           | **90%**   | Asistente RAG contextual           |
| 5  | **PredictionService**       | Gemini + estadística    | 6 endpoints           | **85%**   | Forecast de demanda 30 días        |
| 6  | **AnomalyDetectionService** | Gemini + heurísticas    | 4 endpoints           | **95%**   | Detección de fraude y anomalías    |
| 7  | **ContentGeneratorService** | Gemini 2.0 Flash        | 5 endpoints           | —         | Emails y reportes automáticos      |
| 8  | **CabysClassifierService**  | Gemini + CABYS DB       | 5 endpoints           | **98%**   | Clasificación tributaria CR        |
| 9  | **CreditScoringService**    | Modelo matemático       | 6 endpoints           | **88%**   | Score de crédito 0-100             |
| 10 | **AIServiceInterface**      | Contrato                | —                     | —         | Interface para todos los servicios |

## Ejemplo: Flujo OCR → CAByS → Factura Electrónica

```
1. 📄 Usuario sube factura PDF del proveedor
2. 🤖 OCRService extrae: proveedor, líneas, totales (confianza: 92%)
3. 🏷️ CabysClassifier asigna código tributario a cada línea (confianza: 98%)
4. 📋 Datos pre-llenan formulario de Compra + generan asiento contable
5. ✅ Factura Electrónica lista para envío a Hacienda
```


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://senselab.gitbook.io/senselab-docs/docs/diagrams/03-ciclo-vida-dato-ia.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.