NLP Pipeline Architecture

flowchart TD
    Raw["Raw Text:
Hey, can you show me last quarter sales?"]:::io

    subgraph L1["Character Level - Layer 1: Text Preprocessing"]
        direction LR
        Norm["Text Normalization"]:::l1
        Tok["Tokenization"]:::l1
        Norm --> Tok
    end

    subgraph L2["Word Level - Layer 2: Morphological Analysis"]
        direction LR
        Stem["Stemming"]:::l2
        Lemma["Lemmatization"]:::l2
        Stem --> Lemma
    end

    subgraph L3["Word Level - Layer 3: Syntactic Analysis"]
        direction LR
        POS["Part-of-Speech Tagging"]:::l3
        Dep["Dependency Parsing"]:::l3
        POS --> Dep
    end

    subgraph L4["Sentence Level - Layer 4: Semantic Analysis"]
        direction LR
        NER["Named Entity Recognition"]:::l4
        Coref["Coreference Resolution"]:::l4
        NER --> Coref
    end

    Out["Structured Output:
Ready for intent recognition and query execution"]:::io

    Raw -->|normalized tokens| L1
    L1 -->|root forms| L2
    L2 -->|grammatical tags| L3
    L3 -->|entity relationships| L4
    L4 --> Out

    classDef io fill:#90a4ae,stroke:#37474f,stroke-width:2px,color:#fff,font-size:14px
    classDef l1 fill:#bbdefb,stroke:#1565c0,stroke-width:2px,color:#0d2c54,font-size:15px
    classDef l2 fill:#90caf9,stroke:#1565c0,stroke-width:2px,color:#0d2c54,font-size:15px
    classDef l3 fill:#42a5f5,stroke:#0d47a1,stroke-width:2px,color:#fff,font-size:15px
    classDef l4 fill:#1565c0,stroke:#0d47a1,stroke-width:2px,color:#fff,font-size:15px

    style L1 fill:#e3f2fd,stroke:#64b5f6,stroke-width:1px
    style L2 fill:#e3f2fd,stroke:#42a5f5,stroke-width:1px
    style L3 fill:#e3f2fd,stroke:#1e88e5,stroke-width:1px
    style L4 fill:#e3f2fd,stroke:#1565c0,stroke-width:1px

    linkStyle default stroke:#ff8f00,stroke-width:2.5px,color:#e65100,font-size:13px