Move the estate. Keep control.

Proprietary warehouses lock the door behind you

Teradata, Netezza, legacy on-prem marts, and the proprietary cloud warehouses all share one trait: the data lives in a format only the vendor's engine can read. You pay separately for storage, for compute, and often for the privilege of getting your own data back out. The longer the estate runs, the deeper the dialect-specific SQL, the stored procedures, and the BI semantics burrow into a stack you cannot leave without rewriting everything at once.

Veltheon treats this as an engineering problem, not a re-keying exercise. We move you to an open lakehouse: table data in Apache Iceberg or Delta open table formats, sitting on open file formats (Parquet) in object storage you own, with Postgres as the open, standards-based query and serving layer and open APIs on top. No buyer lock-in — any engine that speaks the open standard can read your tables tomorrow, and leaving costs you a configuration change rather than a rewrite. We produce evidence — convertibility scores, risk flags, and effort estimates — before anyone commits to a budget.

How a Veltheon lakehouse migration runs

1. /01

   Lakehouse Readiness Audit (fixed fee)

   A scoped 2–3 week engagement: we inventory your warehouse — schemas, tables, stored procedures, ETL/ELT jobs, and the BI semantic layer — then run automated analysis to produce a convertibility breakdown, a risk register, and a phased effort estimate. You leave with a plan you could hand to any partner, on a fully open target architecture.

2. /02

   Schema & SQL-dialect analysis

   A senior engineer inventories every table, view, and stored procedure, maps the dependencies between them, and separates true business logic from vendor-specific dialect (proprietary types, hints, and built-ins) and dead objects that should be retired rather than carried forward into the lakehouse. This is read-and-reason engineering work, AI-assisted — we do not run a warehouse-dialect parser or an automated dependency-graph builder, and we would rather tell you that than let you discover it in week three.

3. /03

   Conversion to open formats & open SQL

   Tables are landed as Apache Iceberg / Delta over Parquet; proprietary DDL and dialect SQL are converted to open, standards-based SQL on Postgres. AI can assist a first pass. The scope and effort depend on the estate; a senior engineer reviews correctness and signs off every object that touches business logic.

4. /04

   Minimal-downtime data sync & reconciliation

   A bulk load seeds the lakehouse, then change-data-capture holds it in sync with the legacy warehouse while you test — so cutover is a short, planned window we size up front [est.], not a weekend outage. Row counts, checksums, and reconciliation queries prove parity before anyone signs off. Honest boundary: our CDC engine is relational and fixture-proven, it refuses live connections in code, and it ships no warehouse adapters — so on a warehouse engagement the sync layer is designed and built per estate, not lifted off the shelf.

5. /05

   Cutover, hypercare & handover

   Dress-rehearsed cutover with a documented rollback path, followed by hypercare and a clean handover to your internal team. Because the target is entirely open standards, there is no lock-in to Veltheon as a condition of success — every artifact is yours to run or move.

Why an open lakehouse, not another proprietary stack

• Open table formats — Apache Iceberg and Delta are community standards. Your tables are not held hostage by a single vendor's engine; any compliant query engine can read them.
• Open file formats you own — data lands as Parquet in object storage on infrastructure you control, separating cheap storage from elastic compute instead of metering both together.
• Postgres as the open serving layer — a standards-based SQL engine with a vast ecosystem, no per-core licensing trap, and no proprietary dialect to lock you in next time.
• Designed so leaving is cheap — open formats plus open APIs mean your next migration, if you ever need one, is a configuration change, not a rewrite. Lock-in is the cost we engineer out; we won't pretend the number is literally zero, only that it is a fraction of what a proprietary format imposes.

What a Lakehouse Readiness Audit looks like

Below is a worked example for an illustrative mid-market warehouse estate: 1,800 tables, 620 stored procedures, and 40 ETL jobs feeding a proprietary engine. The audit turns a vague "we need to get off this warehouse" into a costed, risk-rated plan on an open target.

Illustrative only — a real engagement establishes the actual numbers on your data. Table counts, percentages, effort, and risk are examples to show the shape of the output, not a quote.

Engagement & pricing

We start small and de-risk before you commit to the full program. The audit fee is credited against the migration if you proceed with Veltheon — and the deliverable is yours either way, on a fully open target architecture.

We don't publish a figure before we've seen the estate — the fee depends on table count, stored-procedure depth, dialect complexity, pipeline landscape, and BI surface, all established during the audit, and any number quoted ahead of that would be wrong in one direction or the other. Fixed fee in Canadian dollars, in writing, before you commit; see how we scope and quote. Object storage and any managed-Postgres hosting are contracted separately through your cloud provider.

Why Veltheon

• Senior-led, AI-accelerated — a senior data engineer owns every decision; the tooling removes the grind, not the judgment. Veltheon is deliberately small: additional capacity is contracted per engagement, never a standing bench.
• Open standards by default — we target Apache Iceberg / Delta, Parquet, Postgres, and open APIs to reduce switching costs. Remaining platform dependencies are recorded in the decision and handover documents. Aligned with the open lakehouse and PostgreSQL communities; not a certification or endorsement by any vendor.
• Minimal-downtime by design — continuous change-data-capture sync and a rehearsed, reversible cutover, with parity proven against a baseline we lock with you up front.
• Portability by design — the audit deliverable and migration artifacts are yours to take to another qualified partner. Open formats reduce dependency; changing platforms can still require engineering, testing and provider charges.

Apache Iceberg and Apache Parquet are trademarks of the Apache Software Foundation; Delta Lake is a trademark of the Linux Foundation; PostgreSQL is a trademark of the PostgreSQL Community Association of Canada — used for identification only.