BigQuery / Snowflake / Looker Export

The data warehouse export plugin streams a daily delta of your fleet data into the warehouse you already use. It's the standard pick for analytics teams, finance teams, and anyone who's outgrown the built-in dashboard.

What gets exported

Every export run includes new and updated rows from the past 24 hours in these tables:

Each row carries a _exported_at timestamp and _op field (insert, update, delete) so your warehouse can apply changes correctly without full re-loads.

Supported destinations

• Google BigQuery — direct streaming inserts or batch loads from GCS
• Snowflake — Snowpipe streaming or batch loads from S3
• Amazon Redshift — COPY from S3
• Looker Studio — read from BigQuery
• Generic S3 / GCS — Parquet files for self-service ingestion

A single install can target one destination. To export to multiple warehouses, install the plugin once per destination — each counts as a separate $199/month subscription.

Setup

You'll need warehouse credentials with write access to a dataset / schema. Levy creates the tables on its first run.

First-run backfill

The first export includes all of your historical data, not just the last 24 hours. For a fleet with 50,000 rides and 5,000 customers, the backfill typically completes in 30-60 minutes. You can watch progress in Settings → Plugins → BigQuery → Status.

After the backfill, daily incremental exports run at the configured time and complete in a few minutes for most fleets.

Schema stability

The export schema follows semver:

• Patch changes (new optional columns, performance tweaks) deploy silently. Your queries are unaffected.
• Minor changes (new tables) deploy silently. Your existing queries are unaffected.
• Major changes (renamed columns, dropped tables) are announced 90 days in advance via a banner on the installed plugins page and an email to subaccount admins. You can stay on the old schema for up to 6 months while migrating queries.

Levy guarantees zero breaking changes within a minor release line.

Sample queries

Once your data lands in the warehouse, common questions:

-- Daily revenue by subaccount
SELECT
  DATE(end_time) AS ride_date,
  SUM(total_cents) / 100.0 AS gross_revenue_usd
FROM rides
WHERE end_time >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY)
GROUP BY ride_date
ORDER BY ride_date DESC;

-- Vehicles with declining utilization
SELECT
  v.vehicle_number,
  COUNT(r.id) AS rides_last_7d,
  COUNT(r.id) FILTER (WHERE r.end_time &gt;= CURRENT_DATE() - INTERVAL '14 day' AND r.end_time < CURRENT_DATE() - INTERVAL '7 day') AS rides_prior_7d
FROM vehicles v
LEFT JOIN rides r ON r.vehicle_id = v.id
GROUP BY v.vehicle_number
HAVING rides_last_7d < rides_prior_7d * 0.5;

-- Customer LTV cohort
SELECT
  DATE_TRUNC('month', c.created_at) AS cohort_month,
  COUNT(DISTINCT c.id) AS cohort_size,
  SUM(r.total_cents) / 100.0 AS cohort_revenue_usd
FROM customers c
LEFT JOIN rides r ON r.customer_id = c.id
GROUP BY cohort_month
ORDER BY cohort_month DESC;

Pricing

$199/month per destination, billed monthly via Stripe. Levy collects the standard 20% platform fee (this is a first-party plugin, so the fee accrues to Levy directly). No usage-based component — unlimited rows exported.

Free trial: first 30 days are free. You can cancel anytime during the trial with no charge.

Compliance notes

• All exports use TLS 1.2+ in transit.
• Credentials are stored encrypted at rest in Supabase Vault.
• PII export is gated behind the read:customers.pii permission — without it, names, emails, and phones are masked with stable hashes (so you can still join across exports without seeing actual values).
• Levy retains export job logs for 90 days for debugging.
• GDPR right-to-erasure: when a rider deletes their account, Levy emits a delete row in the next export so your warehouse can apply the deletion downstream.

Troubleshooting

What's next

• Configuring a plugin — rotate credentials, change schedule
• Plugin billing — invoicing details