Databricks chargeback built on attribution your teams can audit.
LakeSentry rolls attributed Databricks spend into chargeback: org unit → department → team hierarchies, split and proportional rules for shared infrastructure, and CLI export into billing, finance, or BI systems. Each charged dollar keeps its attribution path and confidence tier, which is what makes the number stand when a team questions it.
Read-only by default. Write access only if you choose to execute actions.
Why Databricks Chargeback Usually Stalls
Chargeback fails on the contested line items, and Databricks produces plenty of them: the all-purpose cluster four teams share, serverless SQL running under one service principal, platform overhead that belongs to nobody’s job. Show a team a charge they can’t verify, and the whole model loses credibility on the spot.
If every workload ran on dedicated, tagged compute, a billing export and a spreadsheet would do. Most environments aren’t that, and the spreadsheet’s owner becomes the arbiter of every dispute.
- Shared compute has no fair split in raw billing exports.
- Overhead needs a standing distribution policy instead of a monthly judgment call.
- Every disputed line lands on the platform team as ad-hoc SQL.
How LakeSentry Handles Chargeback
The same attribution engine, rolled up into a hierarchy finance can bill against.
A hierarchy that matches your org
Org unit → department → team rollups, with projects as cross-cutting labels for context. Teams resolve through identity mappings and rules, so an org change means updating a mapping rather than re-tagging the estate.
Shared costs, split defensibly
Split rules allocate a resource across teams by fixed percentages. Proportional rules distribute platform overhead by each team’s compute-spend share, with exclusions for teams that shouldn’t carry it. Shared clusters split automatically by observed usage.
Export where finance works
Team-level rollups live in the UI, and the LakeSentry CLI exports them into downstream billing, finance, or BI workflows on your schedule. Direct ERP integrations aren’t shipped today; the export is the integration point.
Spreadsheet Chargeback vs LakeSentry
The difference shows up the first time somebody disputes a number.
| Billing export + spreadsheet | LakeSentry | |
|---|---|---|
| Shared clusters | Split by negotiation | Split by observed usage, labeled as estimated |
| Platform overhead | A judgment call each month | Proportional rules by compute-spend share, with exclusions |
| Disputes | Re-derive the number by hand | Every line carries its rule, attribution path, and confidence tier |
| Org changes | Rework the spreadsheet | Update the mapping once; rollups follow the new structure |
| Cadence | A month-end exercise | A continuously maintained ledger — export any time |
Go Deeper
The mechanics live in the docs; the reasoning lives on the blog.
Chargeback runs on the same attribution as everything else — every tier includes it, starting free with three months of history.
Free
$0€0
Standard
$499€499/mo
Pro
$849€849/mo
No per-DBU tax. No per-workspace fees. Paid tiers billed annually.
Frequently Asked Questions
What’s the difference between chargeback and showback?
How are shared clusters charged back fairly?
Can some teams be excluded from overhead allocation?
Does LakeSentry integrate with our ERP or billing system?
What do we show a team that disputes its charge?
See it in your own environment.
The free tier covers unlimited workspaces with three months of history. No card required.