DBPorter
v1.0 · Oracle → PostgreSQLConverts the structure and code of a database from one engine to another. AI adapts the translation to the context instead of applying fixed rules.
Migrating from Oracle to PostgreSQL means rewriting tables, views, functions, and procedures by hand, one object at a time. DBPorter extracts the data definition language (DDL) from the source database, converts it into an abstract intermediate format, and translates it for the target database with AI support. It flags the critical objects that need review in advance. It runs in containers, on-premises or on any cloud, with no vendor lock-in.
Migrate from Oracle to PostgreSQL without rewriting everything by hand
Oracle → PostgreSQL migrations are slow and manual. Every object must be converted, many functions and procedures have no direct equivalent, and rule-based tools break on edge cases. DBPorter is most useful in these situations:
Product specifications
| Product | DBPorter |
|---|---|
| Category | Database migration tool — cross-engine DDL conversion |
| Source (v1.0) | Oracle |
| Target (v1.0) | PostgreSQL |
| Scope (v1.0) | Structure and code conversion (DDL). Data migration roadmap |
| Approach | Abstract intermediate format + AI: context-adaptive translation, not fixed rules |
| Deployment | Containers on Kubernetes (or a Linux VM with Minikube/Kind); on-premises or on any cloud |
| Stack | Fully open source (Java 21/Spring, Liquibase, RabbitMQ, Longhorn S3 storage, and alternatives) |
| Security | Secrets in OpenBao, Keycloak authentication, code scanning with Black Duck |
| Commercial model | On-premises at a fixed cost, with installation managed by Miriade. Pay-per-use on a cloud marketplace coming soon |
From assessment to deployment, through an intermediate format
Assessment
Analyzes the objects to convert and estimates their complexity. It checks them against the rules of the target database and flags critical elements in advance (for example, Oracle functions with no PostgreSQL equivalent).
Extraction
Creates an abstract structure of the source DDL (Liquibase + ANSI SQL), independent of the source engine.
AI-assisted conversion
Translates the source language into the target language with a large language model (LLM). It adapts the translation to the context instead of applying rigid rules.
Deployment
Loads the converted DDL into the target database through dedicated ETL processes, with integration tests at every step.
Database objects, not just tables
In the first release, DBPorter aims to cover at least these object types in Oracle → PostgreSQL conversion:
Adaptive AI, no lock-in
No cloud lock-in
The container architecture lets it run on-premises or on any cloud, with no dependency on a specific vendor.
Adaptive AI translation
It does not apply fixed rules. It uses AI to adapt the translation to the context and the request, and it handles edge cases better.
Abstract intermediate format
Because it works through an intermediate language, the approach is not tied to a single database pair. It is designed to translate across different database types.
Fully open source
Built only on open source technologies, with code security checks (Black Duck) at every stage.
What it takes to run it
Customer-side prerequisites
Security and governance
The roadmap
The first release focuses on converting the structure (DDL) from Oracle to PostgreSQL. The following enhancements are planned but not yet available:
Roadmap — not yet available
After the migration
Database Managed Service (MirDB)
Continuous management of the target PostgreSQL database, for operations and performance after the migration.
Product · Annual managed serviceMirCrypt
Post-quantum encryption of critical data on the migrated PostgreSQL database, with Miriade support.
Platform · Data warehouseDWH4CLICK
Build the enterprise data warehouse on top of the consolidated PostgreSQL database, without writing queries.
Frequently asked questions
Which databases does it support today?
The first release covers conversion from Oracle to PostgreSQL. Support for other database pairs is on the roadmap, enabled by the abstract intermediate format.
Does it also convert data, or only the structure?
In the first release, it converts structure and code (DDL). Data migration, with a validator that checks consistency between source and target, is on the roadmap.
How does it handle Oracle functions with no PostgreSQL equivalent?
It flags them during assessment, together with the possible actions. In the first release, these cases must be rebuilt as custom code or agreed with the customer. Installing Orafce on the customer side reduces these issues.
Where does it run? Am I tied to a specific cloud?
No. It runs in containers on Kubernetes (or on a Linux VM with Minikube/Kind), on-premises or on any cloud. The container architecture prevents vendor lock-in.
Is it an open source product?
It is built entirely on open source technologies. The product itself is a commercial offering from Miriade.
How is it sold?
DBPorter is available on-premises at a fixed cost, with installation managed by Miriade. Pay-per-use availability on a cloud marketplace is coming soon. Contact us to purchase and activate it.
Move your database from Oracle to PostgreSQL
Activate DBPorter in your environment. Tell us your Oracle version, your target PostgreSQL version, and the types and volume of objects to convert. We complete the purchase and size the migration together with you.
Purchase DBPorter for your migration
Tell us about your environment: Oracle version, target PostgreSQL version, types and volume of objects to convert, and on-premises or cloud deployment. We complete the purchase and size the migration together with you.
We use the data you submit only to contact you about this request. Do not enter credentials, passwords, or sensitive data: database access will be handled through agreed secure channels.
