Skip to main content
Schedule a demo

Structural. Measurable. Predictable.

The pricing model you know.A unit you can predict and control.

Buy capacity for the year and draw it down as you build, the way you already buy Snowflake credits.

The difference is the unit. A Complexity Unit measures the schema of your data (how many attributes it has, how deeply they nest, and how you reshape it), not its volume.

Your schema is known before you build. Your volume isn't. Units are used only when you choose to build, so you know your price before you commit, and it doesn't grow with your data volume.

Standards are simpler still: one flat annual fee per X12 transaction set.

See the app in a demo, then your price on your own data in a guided proof of value.

A familiar model.

Annual capacity, drawn down as you build.

A unit you control.

Used only when you activate the build tools.

Volume doesn't count.

Rows, files, and refreshes never use up units.

You'll still pay Snowflake for compute. Scanning and creating reconciliation metrics add a small amount; refreshing the tables costs about the same as any incremental pipeline, and less than full refreshes.

The full cost of your data

The license is the smallest line you can see. The bigger costs are the ones nobody budgets for.

What most budgets track

What they usually don't

Engineering time, technical debt, delay, the cost of data that never arrived, and the value of data you never got to use.

Today

  • Tools and services
  • Compute to refresh pipelinesFull refresh or incremental, depending on how each pipeline is built
  • Engineering to build and maintainMapping every feed, fixing every break
  • Technical debtGrows with every schema change: patches, workarounds, and code nobody wants to touch
  • DelayMonths waiting for usable data
  • Missing data ?Unseen until an audit, a wrong answer, or a missed payment
  • Data you can't use yetSet aside because making it analytics and AI ready costs more than it seemed worth

With DataPancake®

  • DataPancake licenseKnown before you build
  • Scanning and reconciliation metrics
  • Compute to refresh Dynamic TablesIncremental, as your data changes
  • Engineering: minutes to hours
  • Now usable: priced on structure, not volume

Illustrative, not to scale. Your proof of value fills in the real numbers, starting with what your current pipelines miss.

“DataPancake has proven to be a superior tool. Its adoption has led to substantial savings in engineering resources, cloud compute costs, and many other areas.”

— Scott Hamilton, Data Engineering Manager, C.H. Robinson
  • $1M

    in SaaS costs and engineering time savings.

    C.H. Robinson

  • 80%

    lower annual compute cost after moving its pipelines to Snowflake with DataPancake.

    C.H. Robinson

  • ~$600

    in compute to scan approximately 3 billion rows, in about 25 minutes.

    C.H. Robinson

Budget for what you can measure.

Data tools priced on tokens, compute, or rows of data are hard to budget, because those numbers move with your data and you find out what you spent when the bill arrives. A schema is different. You measure it before you build, buy the Complexity Units it needs for the year, and unused units roll over.

Other data toolspriced on tokens, compute, or rows of dataDataPancakepriced on your schema, in Complexity Units
What sets the costHow much data flows and how often it runsHow many attributes, how deeply they nest, and how you reshape it
When you know itWhen the bill arrivesBefore you build
What makes it growMore data, more runs, more queriesOnly a larger schema
Unused capacityDepends on the vendorRolls over

Two ways to price

Every data source is priced one way, never both.

Complexity Units

Structural complexity

DataPancake measures your data's structure (attributes, nesting depth, and transformations) before the build tools are turned on. You buy annual capacity in Complexity Units to cover it.

Complexity Score =

Depth-Weighted Attribute Score

+

Transformation Component Score

Deterministic, visible to you, and measured before activation.

  • More attributes, more data types, and deeper nesting mean a higher score. Data volume never does.
  • The formula is locked for your term.
  • Overage is charged at the same rate as your capacity.
  • Unused Complexity Units roll over.
  • JSON
  • XML
  • Avro
  • Custom data

Flat annual fee

Standards libraries

When your data follows a standard DataPancake has a library for, you pay one flat annual fee per transaction set (for example, 834 enrollment or 837P professional claims). The library comes preloaded, ready to use.

  • X12 834
  • X12 837P
  • One flat annual fee per library. Your data isn't scored or counted.
  • Covers every environment and any number of data sources.
  • Custom fields from trading partners are included.
  • No rollover, no overage.
Which applies to you?

If you apply an X12 library to a data source, that data source is priced by the library. Everything else is priced on structural complexity.

Either way, the Assurance Tier sets the rate.

Regulatory is the standard rate. Enterprise is a discounted rate for data that doesn't carry the same regulatory exposure. Each data source has its own tier, so one organization can use both. Government agency data is always Regulatory.

RegulatoryStandard rate

For data subject to legal, regulatory, audit, government, or operational exposure where structural accuracy carries greater responsibility and consequences of error.

Typical examples

  • Healthcare and FHIR data
  • HIPAA, GDPR, SOX, and similar compliance contexts
  • Government and public sector systems
  • Financial and statutory reporting

EnterpriseDiscounted rate

For business-critical data sources that support production systems, enterprise analytics, AI workflows, and operational reporting without specific regulatory compliance requirements.

Typical examples

  • Customer platforms
  • Operational data systems
  • Enterprise analytics and AI pipelines
  • Large-scale application integrations

From assessment to activation

  1. Connect

    Install DataPancake from the Snowflake Marketplace and connect a data source, in your own account.

  2. Assess

    For data priced on structural complexity, DataPancake measures the schema before any build tools are turned on. Data sources using a standards library skip this step, because the fee is fixed.

  3. Plan

    You see your projected annual price and choose the capacity and libraries you need.

  4. Activate

    Turn on the build tools and generate.

    Units are used here

This assessment is part of the guided proof of value, so you see the price on your own data before you commit.

Included either way

No per-user fees.
No feature gates.

  1. Load

    • Validated data loading and inventory
    • Data loader infrastructure deployed in your account: staging and inventory tables, tasks, views, stored procedures, and functions
  2. Discover

    • Schema discovery across 100% of your data
    • Schema drift detection, regeneration, and complete schema history
    • Alerts when the schema changes
  3. Build

    • Schema metadata configuration
    • Code generation for Dynamic Tables and views
    • Data dictionary and semantic model generation
  4. Prove

    • Reconciliation that proves your tables match the source
    • Results stored in your account

Work your way

The app, script-based bulk updates, and the MCP server for Cortex Code

Buying for commercial enterprises

Buy directly from us, or through Snowflake Marketplace. Either way, the pricing model is the same.

  • Direct from us.An annual order form with TDAA, invoiced by us. Works for any Snowflake customer, whatever your Snowflake contract or region.
  • Through Snowflake Marketplace.Buy DataPancake on Snowflake Marketplace and it's billed through your Snowflake account. If your organization has a Snowflake capacity commitment and is enrolled in Marketplace Capacity Drawdown, the purchase can draw down that commitment, so DataPancake is paid from budget you've already committed to Snowflake instead of a new line item. Your Snowflake account executive can confirm eligibility and enrollment.

Direct from us · pay with

  • Invoice from TDAA

Snowflake Marketplace · pay with

  • Your Snowflake bill
  • Your Snowflake capacity commitment

Buying for government

Agencies can buy DataPancake through Carahsoft and NASPO ValuePoint. DataPancake runs in Snowflake's commercial, commercial government, and government cloud regions.

Buy through

  • Carahsoft
  • NASPO ValuePoint

Runs in Snowflake regions

  • Commercial
  • Commercial government
  • Government cloud

For finance: build your business case

Build your business case with your own numbers.

Your finance team supplies the inputs; we help fill them in during the proof of value.

Use the same rates, volumes, and costs in both columns. Only the hours, months, incidents, and datasets change.

1. Cost savings (annual)

Cost line

  • Today (annual cost)
  • With DataPancake (same measure)
Maintenance and technical debt
Engineer hours per month maintaining these pipelines, including workarounds and rework after schema changes × loaded hourly rate × 12
Engineer hours per month reviewing changes and regenerating × the same rate × 12
Adding new data
Hours to bring a new source into your tables (a new partner, application, system, or file type) × new sources per year × loaded hourly rate
Hours to scan, configure, and generate a new source × the same sources per year × the same rate
Compute
Current annual compute for these pipelines
Projected annual compute from the proof of value: scanning, reconciliation metrics, and incremental refreshes
Certification and audit evidence
Hours per year assembling evidence for audits, examinations, and certifications (sampling, queries, reconciling by hand, answering examiner requests) × loaded hourly rate + outside fees for lookbacks, audit support, or consultants
Hours per year pulling reconciliation results and source-to-column lineage for the same audits × the same rate + outside fees you still need
License
Current tools and services for these pipelines
DataPancake's annual price, plus any current tools you keep

2. Losses avoided (annual)

Loss line

  • Today (annual loss)
  • With DataPancake (same measure)
Missing data
Incidents from missing or wrong data in the last 12 months × cost per incident (rework, penalties, write-offs, missed recoveries)
Incidents still expected once reconciliation checks every load × the same cost per incident (the proof of value shows which past gaps it catches)

Count each regulatory finding or penalty in one row only.

3. New value

Value line

  • Today
  • With DataPancake (same measure)
Data you can't use yetannual
$0 realized from each dataset set aside because making it analytics and AI ready costs too much
For each dataset you now refine, the annual value of the decisions, reports, recoveries, products, or AI use cases it supports (its cost is already in the compute and license lines)
Delayone-time, first year
Months until the data is usable × monthly value of having it
Months until usable with DataPancake, from the proof of value × the same monthly value
Net annual savings
= Today − With DataPancake, across the cost savings rows
Losses avoided
= Today − With DataPancake for missing data
New value
= annual value of data you can now use + one-time value of having it sooner

Judge payback on savings alone. Losses avoided and new value are upside.

Start from your number, not ours. The proof of value runs reconciliation on your own data, so the missing-data line starts from what your current pipelines actually miss.

Build your business case with us

FAQ

How is DataPancake priced?

By the structure of your data (Complexity Units) or by standards library (a flat annual fee per X12 transaction set). Each data source is priced one way, and you know the price before you build.

Why do you price on the schema instead of data volume or usage?

Because the schema is what determines the work. The metadata DataPancake discovers, the configuration you set, and the code it generates all grow with how many attributes and data types your data has, how deeply they nest, and how you reshape it. They don't grow with how many rows flow through. A billion rows with a simple schema is less work to model than a few thousand records with a deeply nested one. The schema is also something you can measure before you build, so your price is known up front. Volume and usage move with your business, and you only see them on the bill.

What counts toward the Complexity Score?

These factors reflect the work required to turn complex semi-structured data into queryable relational tables and views. Attributes: the fields and structures discovered in the source data, counted for each data type. Depth: the nesting level and hierarchy of those attributes; deeper attributes count a little more. Transformations: the configuration required to normalize complex structures into relational form; each adds a set amount. Not counted: virtual attributes you define, and the staging, inventory, and reconciliation tables DataPancake provides.

What does a library fee cover?

One library across development, test, staging, disaster recovery, and production, for any number of data sources that use it. Custom fields your trading partners add are discovered and configured within the same fee, and library updates during your subscription are included. A new version of a standard is a separate library.

Does data volume affect the price?

No. Not rows, not executions, not data volume. Structural complexity is measured from the schema, and library fees are flat.

What happens if our schema grows?

Your score rises only when your schema does, for example when a trading partner adds fields or you configure new transformations. Data volume never raises it. Any overage is charged at the same rate as your capacity, and unused units roll over.

When are Complexity Units used?

Only when you activate the build tools for a data source. The assessment, including measuring your schema and projecting your price, doesn't use any.

What if our X12 data has custom fields from trading partners?

They're discovered and configured within the library fee. A data source with a library applied isn't moved to structural pricing because of custom fields.

Do we pay separately for development, test, and production?

No. A library fee covers every environment. Complexity Units are counted once per data source, not per environment.

What about Snowflake compute?

Snowflake bills compute directly, as it would for any approach. Scanning adds a small amount of compute, and so does creating reconciliation metrics. Refreshing the tables costs about the same as any incremental pipeline, and less than full refreshes.

What happens if we stop using DataPancake?

Everything it generated keeps running. The output is ordinary Snowflake objects with no runtime dependency on DataPancake. If you use X12 libraries, ask us about options for maintaining your X12 license subscription.

Can we see the price before committing?

Yes. The assessment in the guided proof of value shows your projected annual price on your own data.

“With 1.5 engineers in a few weeks, we used DataPancake to generate thousands of lines of production-ready code that would have taken a full team months to write by hand.

Without it, this migration would not have been economically feasible or completed on time.”

— Jon Edwards, VP Carrier Data Acquisition and Integration, Quility

See your price on your own data.

Schedule a demo to see DataPancake in action, then a guided proof of value that measures your data and shows your projected annual price before you commit.