Workload & cost profile
What actually runs, how often, on what volumes, and what it costs today — drawn from query logs and schedules, not from the architecture diagram someone drew two years ago.
The flagship engagement
A fixed-fee, fixed-scope assessment that turns “we should probably fix our data platform” into a costed, sequenced, de-risked plan — whether the answer is a new platform, a migration, a model lifecycle or an agent. No open-ended discovery. No day-rate meter running. A number, a target architecture, and an honest recommendation — in 10 business days.
What you get
Everything below is written down and handed over, whether or not you engage us for the build. The assessment is useful on its own — even if the answer is “not yet.”
What actually runs, how often, on what volumes, and what it costs today — drawn from query logs and schedules, not from the architecture diagram someone drew two years ago.
The Databricks design: Unity Catalog model, compute topology, ingestion pattern, security boundary and network layout, drawn to the level your architects can actually review.
Modeled run cost on Databricks for your real workloads, with the assumptions exposed, so finance is comparing like for like and not a vendor’s best-case slide.
The things most likely to derail the migration — technical and political — ranked, with the mitigation for each. The risky slice we would prove first is named here.
The phased path from here to cutover, with fixed-fee stages, so you can fund the next step without signing a blank six-figure statement of work.
Including “don’t, this year” if that is the right answer. We put it in writing. Our reputation is worth more than one engagement.
The ten days
One senior architect and one senior engineer, focused, for two working weeks. You get a short check-in at the end of each phase, not a black box that emerges with a deck.
Query logs, job schedules, lineage, volumes and spend. We build the real workload profile.
Days 1–4Unity Catalog model, compute, ingestion, security and cost envelope — reviewed with your architects.
Days 5–8We model the cost, rank the risk, and sequence the implementation into fixed-fee phases.
Days 9–10A working session, not a slide read-out. You leave able to defend the plan to your board.
Day 10Where it sits
The assessment is step two. Each step is useful on its own and each one earns the next — you never have to bet big to find out whether we are any good.
Step 01
Sixty minutes with a senior architect. A target-state sketch and the three changes worth making first.
Step 02
The fixed-fee engagement on this page. Profile, design, cost, risk and plan in ten business days.
Step 03
One real workload, end to end, in production-shaped code with reconciliation against the source.
Step 04
The full migration or platform build, in increments, alongside your engineers.
Straight answers
Read access to query logs and schedules, time with the people who know the estate, and a working session or two. We can operate inside your environment — your data does not leave your cloud for us to do this.
Because a day-rate discovery has no incentive to end. Fixed fee means we carry the risk of scoping it wrong, not you, and it forces us to be disciplined about what actually matters in ten days.
Yes. If you engage us for the production build, the assessment fee is credited against it. The assessment is designed to earn the next step, not to be a profit center on its own.
Then we write that down and explain why. Sometimes the licenses do not renew for two years and the smart move is to wait; sometimes the estate needs cleanup first. You still keep every deliverable, and you have saved yourself a far more expensive mistake.
Start here
Bring your current architecture, data volumes and the workloads that hurt. You leave with a target-state diagram on Databricks, the three changes that would cut the most cost or latency, and an honest answer on whether you need us at all.
No slide deck. No procurement process. One senior architect on the call.