Question 1

What is Novuz?

Accepted Answer

Novuz converts legacy data warehouse logic into production-ready cloud pipelines, deterministically, and proves every object. It converts schemas, stored procedures, ETL and orchestration, and runs inside your environment.

Question 2

Is Novuz a tool or a service?

Accepted Answer

Novuz is software, not a services team. Your engineers run it inside your environment. The resolver handles edge cases with engineer approval, and Novuz engineers back every engagement. Partners can run the program around it.

Question 3

How much manual work is left?

Accepted Answer

Typically 95% or more of objects convert with no manual edit. A one-line fix counts as manual, and dead code is excluded from the count. The reconciliation report lists every open item with its cause and fix.

Question 4

What does Novuz migrate?

Accepted Answer

Database schemas, stored procedures and functions, ETL packages, scripts, and workflow and orchestration, plus the data itself. Converted code is native to your target, with no Novuz runtime in it.

Question 5

Which sources and targets are supported?

Accepted Answer

Sources: SQL Server (T-SQL), SSIS, Teradata including BTEQ, Hadoop (Hive, Impala) and Snowflake. Targets: Databricks, Microsoft Fabric, Snowflake, Apache Iceberg and Delta Lake. Not today: Oracle, mainframe, Informatica, DataStage, Netezza and Redshift. Coverage is confirmed per estate during the assessment.

Question 6

Is the conversion AI-based?

Accepted Answer

No. The conversion is rule-based. Fixed, versioned rules, selected by inference over each object’s patterns, convert about 95% of objects with no model calls. The same input gives the same output every time; run it twice and the diff is empty.

Question 7

How is a language model used, and where does my code go?

Accepted Answer

Only the resolver uses a language model, for the roughly 5% of objects fixed rules do not cover. It runs on a model you approve: Claude or Gemini under your own agreement, or a local model through Ollama with no external calls. An engineer approves every proposal. No outbound connection is required; if you choose a hosted model, only the edge-case fragments go to it. There is no shared learning across customers.

Question 8

Does Novuz use AI tokens?

Accepted Answer

Rule-converted objects use none. The resolver uses 700 to 1,000 tokens per edge case. The price is fixed before work starts, so token use never changes what you pay.

Question 9

What results have customers seen?

Accepted Answer

On delivered estates, typically 95% or more of objects convert with no manual edit, and reconciliation is above 95% on first pass. Estates of 1,000 to 1,600 objects take 4 to 8 weeks. Novuz is about 40% cheaper than an SI-led rewrite, at a fixed price set before work starts. Definitions are on the Methodology page.

Question 10

Can Novuz handle mission-critical systems?

Accepted Answer

Yes. Novuz is built for regulated and audit-heavy estates. Every object is checked at schema, row-count and column level and ships with a reconciliation report your auditor can read.

Question 11

Do I need to install anything?

Accepted Answer

Not to start. The scorecard and the assessment need nothing installed; for the assessment you can run the discovery container yourself and send only the report. For the migration, Novuz runs as a Docker container or VM inside your network, air-gapped if you need it.

Question 12

How is Novuz different from platform-native tools?

Accepted Answer

Platform-native tools are a good start for SQL translation to their own platform. Novuz converts the procedures, ETL and orchestration they leave behind, to five targets. It proves every object with a reconciliation report and prices the estate before work starts. Many teams run the free tool first and bring us the remainder.

Question 13

Is Novuz agentic?

Accepted Answer

The conversion is not; it is rule-based and gives the same output every run, with the small share rules cannot resolve approved by an engineer. What is agentic is the access: agents can drive the platform through its APIs and operate the delivered pipelines through MCP.

Question 14

What is logic debt?

Accepted Answer

Logic debt is the business logic locked in legacy stored procedures, ETL packages and schedules. Data migration is solved. Logic migration is not: the logic sets the cost and the risk of a migration.

Question 15

How much does it cost?

Accepted Answer

A fixed price set before work starts, based on object count and complexity, never hours. The estate assessment is $3,000, credited. A First Domain is $25,000 to $40,000, credited. A full estate is fixed from the assessment, typically $40,000 to $200,000 for 1,500 to 10,000 objects.

Question 16

Who is Novuz not for?

Accepted Answer

Estates under about 300 objects; Oracle-only or mainframe estates; teams with no target and no date inside 18 months; buyers who want staff augmentation rather than software; and organizations that cannot run software inside their environment or share code for an assessment.

Question 17

How do I evaluate Novuz?

Accepted Answer

Take the ten-minute logic debt scorecard. Or get the $3,000 estate assessment: a fixed price on your own code in three business days. Or book a call with the people who built it.

Source	To Google BigQuery	To Databricks on Google Cloud
SQL Server (T-SQL procedures, views, functions, SQL Agent jobs)	GoogleSQL and BigQuery stored procedures; cursors restructured as set-based SQL or loops	Databricks SQL or PySpark, and Workflows
SSIS packages	GoogleSQL load and transform steps, with the orchestration agreed in the assessment	Notebooks and Workflows
Teradata (SQL, BTEQ, macros, load utilities)	GoogleSQL and stored procedures, with BTEQ control flow rebuilt as scripted steps	Databricks SQL, notebooks and Workflows
Hadoop (Hive, Impala, Spark scripts, Oozie)	GoogleSQL over BigQuery tables	Databricks SQL, PySpark and Workflows
Snowflake (SQL, JavaScript procedures, tasks, streams)	GoogleSQL and stored procedures, with streams and tasks rebuilt as incremental loads	Databricks SQL, PySpark and Workflows
IBM Netezza (NZSQL, NZPLSQL, nzload)	GoogleSQL and stored procedures, loads from Cloud Storage	Databricks SQL or PySpark, Auto Loader and Workflows
Informatica (IDMC / IICS, Data Quality, Data Engineering Integration)	GoogleSQL pipelines and SQL quality checks	PySpark notebooks, quality checks and Workflows
Mainframe batch (COBOL, copybooks, JCL, DB2, VSAM)	GoogleSQL jobs over tables loaded from the copybook layouts	PySpark jobs and Workflows

Migrate to Google Cloud: BigQuery and Databricks on Google Cloud

BigQuery or Databricks on Google Cloud?

What converts to Google Cloud, by source?

What breaks when logic moves to BigQuery?

What do Google’s own migration tools do, and what do they leave?

How does a migration to Google Cloud with Novuz run?

Where to read next

Frequently asked questions

Sources and further reading

Find out what your logic debt is costing you.