Moon Banking on the Databricks Marketplace

Query Moon Banking data directly inside Databricks. Free preview (top 20 US banks by community vote count) and paid global listing ($1,499/mo) — both delivered as Delta Lake tables with row-level Change Data Feed.

If your team already lives in Databricks, you can pull Moon Banking's bank, vote, story, and aggregate data straight into your workspace as a Unity Catalog share. No API integration, no auth tokens, no rate limits — just a five-table Delta Lake schema sitting next to your other data.

We publish two listings on the Databricks Marketplace:

Moon Banking — Preview — free, instant access, top 20 US banks by community vote count, monthly refresh. Use it to evaluate the schema and data quality before subscribing.
Moon Banking — Full — $1,499 USD / month, every country, daily refresh, row-level Change Data Feed, 7-day time travel.

Tables

Both listings share the same five-table schema. The columns are stable: we commit to additive-only schema changes with 30-day advance notice on any new column, and we never rename, drop, or change the type of an existing column.

Table	What it contains
`world`	Singleton row of global aggregates: country count, bank count, story count, plus the 14-category vote totals and scores.
`countries`	One row per country: name, ISO code, flag emoji, world rank, and per-country aggregate scores across the 14 categories.
`banks`	One row per bank in each country: name, URL, description, world + country rank, story count, and per-bank aggregate scores.
`bank_votes`	Individual upvote / downvote events on a specific bank in a specific category. PII (IP addresses, user IDs) is excluded.
`stories`	User-submitted short reviews about a bank, with sentiment + tags. PII excluded.

The 14 voting categories — crypto_friendly, customer_service, fees_pricing, digital_experience, security_trust, account_features, branch_atm_access, international_banking, business_banking, processing_speed, transparency, innovation, investment_services, lending — are stored as STRUCT<total: INT, up: INT, down: INT, score: INT> columns on world, countries, and banks.

We deliberately strip every form of PII before publishing:

No user identities, names, emails, or accounts
No IP addresses on votes or stories
No internal API keys, fingerprints, or tracking signals
No moderation flags or admin-only fields

Free preview listing

The preview listing is instantly accessible — there's no request flow, no email exchange, no payment. Click "Get instant access" on the listing page in Databricks and the share appears in your workspace.

What's different from the full listing:

Coverage: The 20 US banks with the highest community vote count, refreshed monthly. countries contains only the US row; banks contains those 20 institutions; bank_votes and stories contain every event attached to them. The world row is global aggregates and is identical to the paid listing.
Refresh cadence: Monthly, on the 1st of each month at 05:00 UTC. The 20-bank roster is also recomputed on each refresh, so a bank that climbs into the top 20 will appear in the next month's snapshot.

What's the same:

Identical five-table schema and column docs
Delta Change Data Feed enabled
7-day time-travel retention
Schema-stability commitment

Paid full listing

$1,499 USD per month, billed monthly via Stripe. To subscribe:

Find "Moon Banking — Full" in the Databricks Marketplace.
Click Request access. Databricks asks for your email and (optionally) your workspace metastore ID.
Within a minute or two, you'll receive an email from support@e.moonbanking.com with a Stripe Payment Link.
Pay through Stripe. Our webhook adds your Databricks recipient to the moonbanking_full_share automatically; the listing flips to "Installed" in your workspace.
Open the bundled full_starter.py notebook for snapshot queries, time-travel reads, Change Data Feed sync, and cross-country aggregates.

Cancel any time. Access remains until the end of the current billing period; at that point your recipient is removed from the share automatically. There are no refunds — see the refund policy.

Engineering features

Format. Delta Lake on Unity Catalog. Standard Databricks-to-Databricks share — no client setup required. Open Sharing recipients are also supported for consumers without a Databricks workspace.
Schema. Stable, additive-only.
Time travel. 7 days. Use TIMESTAMP AS OF for backtesting and rollback.
Change Data Feed. Enabled on every table. Use table_changes() for cheap incremental sync into your own Delta tables.
Refresh cadence. Daily for the full listing; monthly for the preview.

Sample queries

Once the share is installed in your workspace, the simplest way to get going is to open the bundled starter notebook (preview_starter.py or full_starter.py). A couple of representative snippets:

-- The 20-bank preview roster, ranked by community score
SELECT name, overall.total AS total_votes, overall.score AS score, rank
FROM moonbanking_preview.preview.banks
ORDER BY overall.score DESC

-- Incremental sync via Change Data Feed (full listing)
SELECT *
FROM table_changes(
  'moonbanking_full.public.banks',
  current_timestamp() - INTERVAL 24 HOURS,
  current_timestamp()
)
WHERE _change_type IN ('insert', 'update_postimage', 'delete')

License

Use of the Moon Banking dataset on the Databricks Marketplace is governed by the Marketplace Data License. Highlights:

Analytics, AI/ML model training (including generative AI, large language models, agents, and other foundation models), fine-tuning, evaluation, research, and dashboarding are permitted. Models you train on the data are yours to ship as derivative works.
Redistribution of the raw dataset, repackaging, exposing the data through a public API, or building a product that is substantially similar in purpose to Moon Banking is not permitted.
See the license for the full terms.

Support

Questions about billing, access, schema, or data quality? Email support@e.moonbanking.com.