Skip to content
Free through Q3 2026

Send us your cloud bill.
Free 90-min cost review.

A live 90-minute working session on your AWS, GCP, Azure, Snowflake, BigQuery, Databricks — or any combination of those. We get on the call with you, share screens, walk through the bill together.

You leave with named levers, a savings model with effort estimates per lever, and a 30/60/90-day plan your engineering team can run with next week. Recording optional.

Why free, and why now. Replatform is in the early-customer phase. The 90-min session runs at no charge through end of Q3 2026 — in exchange for permission to publish the anonymized methodology. Your numbers, names, and topology stay private. After that window, the same session reverts to a 5-day $100 paid diagnostic credited to any execution engagement signed within 60 days.

What you leave the call with

  • Top waste sources, named and ranked by impact in USD/INR with effort estimate per lever.
  • Slot or warehouse rightsizing model with the recommended commit shape, term, and a sensitivity table.
  • Query-pattern teardown — the most expensive query patterns we surface in your environment, with the rewrite for each.
  • dbt model fan-out audit — which models re-run unchanged source data, which multiplicatively explode source-table reads, which nobody queries downstream.
  • 30/60/90-day execution plan separating quick wins (week 1) from architectural changes (quarter 2+), with savings booked against each lever.
  • Monitoring blueprint — what to watch, where to put alerts, how to keep cost from regressing after the work ships.

How the session runs

Before the call. You give us read-only metadata access on the warehouse — query history, slot or warehouse reservations, dbt manifest if you run dbt. No production data access.

During the 90 minutes. We run our codified rule library and deterministic validation framework live against the metadata, with your platform lead on screen-share. We walk through findings as they surface — named levers, the rightsizing model, the dbt fan-out shape, the 30/60/90-day plan. You ask questions in real time. Recording optional.

After the call. We hand your engineering team the plan and step away. Most teams ship the quick-wins themselves within 30 days; the architectural changes either become a follow-on Execution engagement (fee credited if applicable) or land on your roadmap. We don't move money around for you, change your queries, or sell you a SaaS subscription.

The kind of teardown we run

Every Cost X-Ray is methodology-first. The same rule library and validation framework runs against your environment — but the specific levers, magnitudes, and recommended plan are entirely environment-dependent.

We've shipped this methodology on a recent India D2C consumer commerce platform's BigQuery estate, where the engagement codified the kind of waste patterns described above. The actual numbers on your environment will look different.

The sample deliverable (linked at the top of this page) shows the output shape — the 12 sections of the deck, the structural promise of what you'll receive. Treat the numbers in the sample as illustrative of the format, not as a prediction of your outcome. If the X-Ray surfaces $10k of annual waste in your environment, the deck reflects that. If it surfaces $500k, same deck, different numbers.

What the X-Ray is NOT

  • Not a SaaS dashboard. No subscription. No login. A deck, a sheet, and a meeting.
  • Not a 10-day FinOps Audit. The FinOps Audit covers your entire cloud spend — compute, storage, egress, Kubernetes, warehouse. The Cost X-Ray is the warehouse-only teardown at twice the depth. FinOps Audit lives here if you want the broader cut.
  • Not an Execution engagement. If the X-Ray surfaces a rewrite worth shipping, that is a separate $50–300k fixed-fee Execution scoped from the X-Ray output. The X-Ray fee — once reinstated — is credited.
  • Not a guaranteed savings number. We commit to a structured deliverable: top waste sources ranked, slot/warehouse rightsizing model, 30/60/90-day plan. We don't commit to a percentage — the rule library finds what's in your environment, not what's in someone else's.

Pricing & scope

  • Free through end of Q3 2026. The 90-min session runs at no charge in exchange for permission to publish the anonymized methodology — your numbers, names, and topology stay private.
  • After Q3 2026: reverts to a 5-day $100 paid diagnostic, credited 100% to any Execution engagement signed within 60 days of deck delivery.
  • Read-only metadata access only. We do not require production data access.
  • NDA on request. Standard MSA + DPA available — typically a 24-hour legal review on our side.
  • The 1-page credential checklist is shared on the scoping call, scoped to your specific platforms.

Who it's for

$50M–$1B revenue companies running BigQuery, Snowflake, Databricks, or significant AWS/GCP/Azure spend at production scale — typically post-growth, starting to see cloud or warehouse spend pull ahead of the platform team's attention budget.

The buyer is typically a VP of Data, Head of Platform, CTO, or CFO running an efficiency mandate. If the 90-min cost review is the wrong shape for your situation, we'll say so on the first scoping call — and point you at whatever is the right shape.

Questions buyers actually ask

Before you book a call.

How is the X-Ray different from a vendor calculator like the BigQuery slot estimator or Snowflake’s cost insights?
Vendor calculators model one dimension well — slot sizing, credit attribution, storage tiering. They do not name the five most expensive query patterns in your environment, audit dbt model fan-out, or sequence findings into a 30/60/90-day plan with effort estimates per lever. The X-Ray combines the vendor telemetry with a codified rule library built from prior warehouse engagements (including an India D2C platform’s BigQuery estate) and outputs a deck your engineering team can execute against next week, not a dashboard you have to interpret.
What read-only credentials do you actually need? Can our security team review the access scope?
Read-only metadata access on the warehouse — INFORMATION_SCHEMA / ACCOUNT_USAGE / system tables, query history, slot or warehouse reservations, and dbt manifest if you run dbt. No production data access required. The credential checklist is shared at SOW signature so your security team reviews the exact scope before anything is provisioned. NDA on request, standard MSA + DPA available.
What happens to the $100 if we don’t go forward with execution work afterward?
The deck is yours either way. No follow-on ask. The fee is credited 100% to any Execution engagement signed within 60 days of deck delivery — but there is no obligation to engage. Most teams ship the quick-wins themselves; the architectural changes either land on your roadmap or become a scoped follow-on. We do not pursue the conversation if the X-Ray surfaces nothing worth executing on.
What does day 1 versus day 5 actually look like?
Day 1: read-only credentials land, metadata ingestion begins, kickoff call with your platform lead. Days 2 – 4: the rule library runs against your estate; the team works through query-pattern teardown, slot or warehouse rightsizing modelling, dbt model fan-out audit, and waste-source ranking. Day 5: a 90-minute live deliverable walkthrough — recorded — with the 12-section deck, savings model, and 30/60/90-day plan delivered before the session ends.
Is this just FinOps Foundation framework consulting in a wrapper, or do you produce a specific deliverable?
A specific deliverable. The FinOps Foundation framework is one input — useful for the allocation / showback dimension — but the X-Ray output is a 12-section deck with named query patterns, a slot or warehouse rightsizing model with a sensitivity table, a dbt model fan-out audit, and a sequenced execution plan with effort estimates per lever. There is no certification track, no maturity model, no posters. A sample deliverable is linked at the top of this page so you can see the shape before paying.
What if our warehouse is on Redshift or Synapse, not Snowflake, BigQuery, or Databricks?
Redshift is in scope — the founder has shipped Redshift → Snowflake and Redshift → BigQuery migrations, so the rule library covers the source side. The X-Ray on Redshift focuses on WLM queue economics, distkey / sortkey waste, and Spectrum read patterns. Synapse is bespoke; we will say so on the 15-minute scoping call and quote separately if we take it.
Do you sign an NDA before the X-Ray?
Yes. Standard mutual NDA on request before any metadata exchange. MSA + DPA are also available — typically a 24-hour legal review on our side. The intake form does not require any sensitive information; the credential exchange happens after SOW + NDA sign.
Free resource

Not ready for a call?

If you are still scoping whether a migration makes sense, start with the Migration Readiness Checklist — 20 items across all six pipeline stages, with a score rubric that tells you whether you are ready to execute, need to close some blockers first, or need a Discovery Sprint before committing budget.

Download the checklist
How to book

Two steps. 90 minutes on the call.

  1. 1. Send the intake. Tell us which warehouse, rough monthly spend band, and one sentence about what's prompting the look. We reply within one business day with a 15-min scoping call slot.
  2. 2. Pick a time + share read-only credentials. Read-only metadata access on the warehouse only. We come into the 90-min call with the prep done; you leave it with the named levers, savings model, and 30/60/90-day plan.
What it costs

Free through end of Q3 2026, in exchange for permission to publish the anonymized methodology — your numbers, names, and topology stay private. After that window, the same session reverts to a 5-day $100 paid diagnostic, credited 100% against any execution engagement signed within 60 days.