In product
BigQuery · Cost Governance

Rank waste, not spend.
Slots, scans, owners.

Anavsan is the Workload Governance Platform for Snowflake and BigQuery. Powered by APEX, it traces waste (not spend) to a workload and owner, assigns the fix, enforces a decision through your review process, and proves closure from the data.

On-demand bills bytes processed (TiB). Capacity bills slots — reservations, editions (Standard, Enterprise, Enterprise Plus), baseline versus autoscale. Slots are billed per second, with a one-minute minimum. We rank idle baseline, overshoot, and unpruned scan — not the busiest project.

Book a Demo Start Free Trial
The problem

Billing shows the invoice.
It does not rank waste.

Google already totals on-demand TiB and reservation slot-hours. That is spend. Cost governance starts when someone can name which jobs and reservations are waste — and who owns the fix.

Idle baseline and autoscale overshoot

Committed baseline slots sit unused. Autoscale adds capacity that outlives the burst. Editions (Standard, Enterprise, Enterprise Plus) bill slots per second, with a one-minute minimum.

Unpruned scans and SELECT * bytes billed

On-demand charges bytes processed. Partition and cluster pruning that never fires, plus SELECT *, inflate TiB without looking like a “slot problem.”

Unowned jobs

INFORMATION_SCHEMA.JOBS can show user_email and destination table. If that does not resolve to a workload owner, the finding stays marked as needs owner — not assigned to a guess.

How APEX applies

Trace waste to a job. Close it from the data.

Same loop as Snowflake, in BigQuery objects: jobs, slots, reservations, scan. Metadata only. The Private Knowledge Graph remembers ownership — it does not AutoApply.

✓
Rank waste, not project spend
✓
On-demand bytes billed and capacity slot-ms
✓
Named owner — or needs owner
✓
Decision through your review process
✓
Closure when the metric recovers
How APEX applies

Trace → Assign → Enforce → Prove

Four stages. BigQuery vocabulary. No invented signal catalog. No AutoApply.

01 · Trace

Waste on jobs & slots

Rank idle baseline slots, autoscale overshoot, unpruned scans, and SELECT * bytes billed — not the project with the largest invoice line.

02 · Assign

A named owner

Resolve the job or reservation from evidence. Unowned jobs stay needs owner. The Private Knowledge Graph is ownership memory — it does not AutoApply.

03 · Enforce

Via review / PR

The exact change — query, reservation, or job config — goes through your review process until it is applied, deferred with a reason, or dismissed.

04 · Prove

Metric recovered

The finding closes when slot-ms or scanned bytes return to normal. Closure is from the data, not a checkbox.

Honest vs Google native

Google already shows X.
Anavsan closes Y.

We do not replace Cloud Billing or reservation reports. We attach waste to an owner and a closed decision.

Google already shows

Spend, slots, and job stats

  • Cloud Billing export and on-demand bytes billed / TiB
  • Reservations, editions, baseline vs autoscale utilization
  • INFORMATION_SCHEMA.JOBS / JOBS_BY_PROJECT: slot-ms, bytes billed, user_email
  • Capacity Troubleshooter and slot estimator views
Anavsan closes

Waste, owner, review, proof

  • Rank addressable waste — idle baseline, overshoot, unpruned scan — not spend
  • Assign a named owner, or leave the finding as needs owner
  • Enforce the change through your review process — never AutoApply
  • Prove closure when slot-ms or scanned bytes recover
FAQ

Cost governance questions

In product. Metadata only. Waste, not spend.

Does Anavsan rank BigQuery spend or waste?
Waste. Cloud Billing and reservation reports already show spend. Anavsan ranks addressable waste: idle baseline slots, autoscale overshoot, unpruned scans, SELECT * bytes billed, and unowned jobs.
Does this cover on-demand and capacity?
Yes. On-demand is billed on bytes processed (TiB). Capacity is slots, reservations, and editions (Standard, Enterprise, Enterprise Plus), with baseline versus autoscale. Slots are billed per second, with a one-minute minimum.
Does Anavsan read BigQuery table contents?
No — operational metadata and INFORMATION_SCHEMA-class signals only. Anavsan never reads, copies, or moves table contents.
Does APEX AutoApply reservation or query changes?
No. The Private Knowledge Graph is ownership memory for assignment and closure. Changes go through your review process. Anavsan never AutoApplies. The engine write-up lives on APEX.
Is dataset lineage part of cost governance?
No. Cost governance ranks slot and scan waste. Dataset lineage maps JOBS referenced_tables to a named owner and flags unreferenced tables. Same APEX loop; different object. Storage is on storage intelligence.

See waste ranked on
your project.

BigQuery cost governance is in product. Book a 30-minute demo or start a trial — Trace → Assign → Enforce → Prove, on slots and jobs.

Book a Demo Start Free Trial
BigQuery hub → Storage Intelligence → Dataset lineage → APEX engine → Snowflake cost counterpart →