Connect Snowflake to VectorFin Financial Iceberg Data

CREATE OR REPLACE EXTERNAL VOLUME vf_gcs_volume
  STORAGE_LOCATIONS = (
    (
      NAME = 'vf-gcs-us'
      STORAGE_PROVIDER = 'GCS'
      STORAGE_BASE_URL = 'gcs://vectorfinancials-data/warehouse/'
    )
  );

-- Get the GCS service account to grant access
DESC EXTERNAL VOLUME vf_gcs_volume;

CREATE OR REPLACE CATALOG INTEGRATION vectorfin_polaris
  CATALOG_SOURCE = POLARIS
  TABLE_FORMAT = ICEBERG
  CATALOG_NAMESPACE = 'vectorfin'
  REST_CONFIG = (
    CATALOG_URI = 'https://catalog.vectorfinancials.com'
    WAREHOUSE = 'vectorfin_warehouse'
  )
  REST_AUTHENTICATION = (
    TYPE = OAUTH
    OAUTH_CLIENT_ID = '<your_client_id>'
    OAUTH_CLIENT_SECRET = '<your_client_secret>'
    OAUTH_ALLOWED_SCOPES = ('PRINCIPAL_ROLE:ALL')
  )
  ENABLED = TRUE;

-- Transcript embeddings
CREATE OR REPLACE ICEBERG TABLE vf_transcripts
  CATALOG = 'vectorfin_polaris'
  EXTERNAL_VOLUME = 'vf_gcs_volume'
  CATALOG_TABLE_NAME = 'embeddings.transcripts';

-- Quant signals
CREATE OR REPLACE ICEBERG TABLE vf_whystock_score
  CATALOG = 'vectorfin_polaris'
  EXTERNAL_VOLUME = 'vf_gcs_volume'
  CATALOG_TABLE_NAME = 'signals.whystock_score';

-- Or query directly without catalog (GCS path)
CREATE OR REPLACE ICEBERG TABLE vf_signals_direct
  BASE_LOCATION = 'warehouse/signals/whystock_score/'
  EXTERNAL_VOLUME = 'vf_gcs_volume';

-- Top signals by quant score
SELECT ticker, date, score,
       components:regime_score::FLOAT as regime,
       components:volatility_score::FLOAT as vol_score
FROM vf_whystock_score
WHERE date >= '2024-01-01'
ORDER BY score DESC
LIMIT 20;

-- NVDA earnings embeddings
SELECT ticker, fiscal_period, chunk_idx,
       VECTOR_COSINE_SIMILARITY(embedding, query_vector) as similarity
FROM vf_transcripts
WHERE ticker = 'NVDA'
ORDER BY similarity DESC
LIMIT 5;