Database Schema¶

Entity Relationship Overview¶

erDiagram
    COMPANIES {
        int id PK
        string name
        string ticker
        string exchange
        string sector
    }
    FILINGS {
        int id PK
        int company_id FK
        string report_type
        string period
        date filing_date
        string source_url
        string pdf_path
    }
    INCOME_STATEMENTS {
        int id PK
        int filing_id FK
        decimal revenue
        decimal gross_profit
        decimal operating_income
        decimal net_income
        decimal eps
    }
    BALANCE_SHEETS {
        int id PK
        int filing_id FK
        decimal total_assets
        decimal total_liabilities
        decimal total_equity
        decimal cash_and_equivalents
    }
    CASH_FLOWS {
        int id PK
        int filing_id FK
        decimal operating_cf
        decimal investing_cf
        decimal financing_cf
        decimal free_cash_flow
    }
    DOCUMENT_CHUNKS {
        int id PK
        int filing_id FK
        text content
        int chunk_index
        vector embedding
    }
    USERS {
        int id PK
        string email
        string hashed_password
        string tier
        datetime created_at
    }

    COMPANIES ||--o{ FILINGS : has
    FILINGS ||--o| INCOME_STATEMENTS : contains
    FILINGS ||--o| BALANCE_SHEETS : contains
    FILINGS ||--o| CASH_FLOWS : contains
    FILINGS ||--o{ DOCUMENT_CHUNKS : split_into

Tables¶

companies¶

filings¶

income_statements¶

balance_sheets¶

cash_flows¶

document_chunks¶

users¶

Migrations¶

pgvector Setup¶

CREATE EXTENSION IF NOT EXISTS vector;

CREATE INDEX ON document_chunks
USING hnsw (embedding vector_cosine_ops)
WITH (m = 16, ef_construction = 64);