feat(analytics): add Santander bank analytics database and queries

Summary

This PR adds a SQLite database with customer and transaction data, along with analytical SQL queries and a data loading script. The database contains 500 customers and 5,000 transactions loaded from Excel source files.

Files added:

• data/santander_bank.db - Pre-populated SQLite database
• scripts/load_data.py - Python script to recreate the database from Excel files
• scripts/query_1_customer_demographics.sql - Customer count, avg age, avg balance by segment
• scripts/query_2_monthly_transactions.sql - Monthly transaction volume by category
• scripts/query_3_top_customers.sql - Top 10 customers by transaction count
• scripts/query_4_channel_analysis.sql - Average transaction amount by channel

Review & Testing Checklist for Human

• Verify Excel source files exist: The load_data.py script expects santander_customers.xlsx and santander_transactions.xlsx in the data/ directory. These are not committed - confirm if they should be added or documented.
• Test SQL queries against the database: Run each query file against santander_bank.db to verify they return expected results
• Verify database schema: Check that data types (TEXT for dates, INTEGER for booleans) are appropriate for your use case

Recommended test plan:

# Test the queries work
python -c "import sqlite3; conn = sqlite3.connect('data/santander_bank.db'); print(conn.execute('SELECT COUNT(*) FROM customers').fetchone())"
# Expected: (500,)

Notes

• The SQLite database is committed as a binary file (~247KB). Consider if this should be gitignored and regenerated via the load script instead.
• SQL queries use SQLite-specific strftime() function - not portable to other databases.
• Foreign key constraints are defined but not enforced by default in SQLite.

Link to Devin run: https://app.devin.ai/sessions/f8ee95a0e1cd404a86aadd93b70546eb
Requested by: Joao Esteves (@joao-cognition)