The Automated Replication Pipeline Designer is an interactive Gio™ AI Assistant in the Data Integrity Suite that streamlines the process of mapping data between source and target systems. It guides users through setup, connection selection, and table mapping by reducing manual effort and minimizing errors. It is ideal for both new and experienced users configuring data replication.
Key benefits
-
Step-by-step guidance for mapping configuration: The Gio™ AI Assistant walks you through each stage of the setup, so you never have to figure out the sequence on your own.
-
Automated suggestions for field mapping and transformations: The system identifies and pre-fills relevant tables and schemas based on your intent, saving you from manual lookup.
-
Real-time validation and feedback: Errors and missing information are caught immediately, so you can correct them before saving.
-
Consistent, repeatable mapping processes: Every pipeline follows the same structured workflow, making it easier to maintain and audit.
-
Significantly reduced manual setup and configuration effort: What might take hours of manual configuration can be done in minutes through guided conversation.
How the automated replication pipeline designer works
The Automated Replication Pipeline Designer uses the Gio™ AI Assistant to understand what you want to replicate and create the pipeline for you. You can describe your replication goal in one message or build it up gradually. Gio™ AI Assistant will guide you through each step, ask for anything that is missing, and validate your setup before moving forward. Every pipeline is created through three phases that always run in order:
Phase 1: Project and Pipeline Setup
Gio™ AI Assistant creates a new project (or lets you select an existing one) and initializes a pipeline inside it. A project is simply a container that holds one or more pipelines. A pipeline cannot exist without a parent project.
Phase 2: Connection Validation and Linking
Gio™ AI Assistant checks that your source and target database connections are compatible with each other based upon your request, then links them to the pipeline. If the requested connection is not compatible, Gio™ AI Assistant will notify you that the connection request is invalid.
Phase 3: Table and Schema Mapping
Gio™ AI Assistant discovers available tables from your source connection based on the names or patterns you provide and configures the replication mappings. You review the matched tables before the pipeline is finalized and executed. Gio™ AI Assistant handles the sequence automatically and will only ask you for input when it needs to.
What you can do with Gio™ AI Assistant
-
Create a new replication pipeline from scratch: Provide your source, target, and table details all at once or step by step, and Gio™ AI Assistant builds the full pipeline for you.
-
Create a pipeline with just source or target details: You don't need to specify both sides upfront. If you provide only the source, Gio™ AI Assistant fetches available target connections for you to choose from, and vice versa.
-
Create a pipeline with only connection types: If you say something like "Postgres to Snowflake", Gio™ AI Assistant will find all matching connections in your workspace and let you pick.
-
Create a pipeline with no connection details at all: If you just say "Create a replication pipeline", Gio™ AI Assistant will list all available connections in your workspace for you to select from.
-
Browse and select from existing projects: You can ask Gio™ AI Assistant to list all your existing projects and choose one to use for your new pipeline.
-
Create a pipeline under a named new project: You can specify a new project name and Gio™ AI Assistant will fetch the list of existing projects for you to review. To create the new project, select the Create New Project button and Gio™ AI Assistant will create the project with the name you provided and initialize the pipeline inside it. If a project with that name already exists in your workspace, Gio™ AI Assistant handles the conflict automatically by appending a unique identifier to the name, you will not be asked about it.
-
Select tables and schemas using names or patterns: You can specify exact table names, schema names, prefix patterns, suffix patterns, or keyword-based selections in plain language.
-
Add more tables to an existing selection: After reviewing the matched tables, you can ask Gio to include additional tables and it will merge them with the existing selection.
-
Replicate all tables in a schema: You can ask Gio™ AI Assistant to replicate all tables within a specific named schema or select all tables and all schemas at once using clear all-grant phrases.
What you cannot do with Gio™ AI Assistant
-
Modify or edit a pipeline: Once a pipeline has been fully configured and saved, Gio™ AI Assistant cannot reopen or edit it. For changes to a completed pipeline, use the manual configuration options in the Data Integrity Suite.
-
Filter or search projects by name pattern: You can only list all projects. Gio™ AI Assistant does not support filtering projects by name patterns or wildcards such as "Show projects starting with PRODUCTS".
-
List connections without a pipeline context: Gio™ AI Assistant cannot browse or list connections independently without first initializing a project and pipeline. Asking to list connections on their own is outside the scope of this agent.
-
Use vague or placeholder table inputs: Inputs like "all data", "replicate everything", "sync all datasets", or placeholder text like [Tables] or [Schemas] are not accepted. You must provide specific table names, schema names, or patterns.
-
Create custom table mappings: Mapping a source table to a differently named target table by natural language is not supported. Target tables are always created automatically by the system with the same name as the source.
-
Use unsupported connection type combinations: Not all source and target combinations are supported. If you specify an incompatible pair, Gio™ AI Assistant will stop and show you the valid combinations for your source type.
-
Create pipelines for general data tasks: Gio™ AI Assistant is a specialized replication pipeline agent. Requests unrelated to creating or configuring a replication pipeline are outside its scope.
Common use cases
1. Provide Complete Replication Intent
If you already know your source system, target system, and what data you want to replicate, you can provide all of that in a single message. For example:
"Create a replication pipeline from my Oracle database to Snowflake by selecting the "ORDERS_" and "CUSTOMERS_" tables from the "LOGISTICS" schema"
When you do this, the system automatically:
-
Identifies the right replication workflow for your source and target combination
-
Creates a new project and pipeline with the required settings if user opted for creating a new project
-
Discovers and selects the relevant tables and schemas for replication
-
Generates table mappings and replication rules
Once the pipeline is built, review the structure and mappings, and save it for execution or further adjustments.
2. Incremental Intent and AI Refinement
You don't have to have everything figured out upfront. You can start with a partial request, for example, just naming your source system and Gio™ AI Assistant will ask you follow-up questions to fill in the gaps.
"Help me set up a replication pipeline from Oracle."
Gio™ AI Assistant will let you choose the compatible target system, asks for the tables or schemas to replicate, and any other information needed. As you answer:
-
The pipeline is built and updated step by step
-
Tables and schemas are discovered and populated automatically
-
Mappings are done as you refine your requirements
Once all required details are in place, review and save the completed pipeline.
3. Source Table Selection and Mapping
Once your source and target connections are confirmed, Gio™ AI Assistant will ask you to specify which tables or schemas you want to replicate. You can describe this in plain language and Gio™ AI Assistant understands a wide range of inputs:
-
A specific table name: "CUSTOMERS table" or "ORDERS"
-
A specific schema name: "SALES schema"
-
Tables starting with a keyword: "tables starting with CUST"
-
Tables ending with a keyword: "tables ending with _DATA"
-
Tables containing a keyword: "tables with claim in the name"
-
All tables within a named schema: "all tables from the SALES schema"
-
All tables and all schemas: "select all tables and all schemas"
Once you provide your selection, Gio™ AI Assistant fetches the matching tables from your source connection and displays them along with their schema information for you to review. You can confirm the list or request a change before anything is saved.
After you confirm, Gio™ AI Assistant applies the table mappings to the pipeline and sets up the connection-specific replication components. Target datasets are created automatically where supported. Once this step is complete, the pipeline is fully configured and ready for execution.
4. Confirm and Finalize the Replication Pipeline
Before execution, the Gio™ AI Assistant summarizes the planned pipeline including selected systems, objects, and configurations. Review and confirm the summary, then let the AI finalize and save the pipeline.
5. Resolve Ambiguities and Constraints
If your replication intent is ambiguous or contains conflicting information, the Gio™ AI Assistant will pause pipeline creation and prompt for clarifications. Once all ambiguities are resolved, the AI generates the pipeline according to your validated intent.
Typical Workflow
-
Start the Automated Replication Pipeline Designer in the Data Integrity Suite.
-
Provide your replication intent.
-
Respond to AI prompts for any missing or ambiguous information.
-
Review the automatically generated or updated pipeline, including object selections and mappings.
-
Confirm and save the pipeline for execution or further customization.
Sample Queries
Below are example queries you can use to interact with Gio™ AI Assistant. You do not need to follow a strict format, plain natural language also works. The more specific you are, the faster Gio™ AI Assistant can build your pipeline.
Creating a new pipeline with full details provided
-
"Create a replication by creating a new project called "Products" project pipeline from LTIAS05_NoVault Db2i to snowflake-hashicorp Snowflake selecting all the tables under the "PRODUCTS" schema"
-
"Replicate from MainDB Postgres to DataWarehouse Snowflake."
-
"Create a replication pipeline from Oracle to Bigquery."
-
"Postgres to Bigquery replication pipeline."
Creating a pipeline with only source or target provided
-
"Create a pipeline from MainDB Postgres."
-
"Create a replication pipeline from Novault."
-
"Create a replication pipeline from MSSQL."
-
"Create a pipeline to KafkaTopic Kafka."
-
"Create a replication pipeline to SnowflakeWarehouse."
-
"Create a replication pipeline to Snowflake."
Gio™ AI Assistant will automatically find the source connection and fetch the compatible target connections for you to choose from, or vice versa.
Creating a pipeline with no connection details yet
-
"Create a replication pipeline."
-
"Help me set up a replication pipeline"
Gio™ AI Assistant will fetch all available and compatible source and target connections in your workspace and ask you to select a source and target.
Specifying tables and schemas after connections are confirmed
-
"Replicate the CUSTOMERS and ORDERS tables."
-
"SALES schema."
-
"Tables starting with CUST."
-
"Tables ending with _DATA."
-
"All tables from the SALES schema."
-
"Select all tables and all schemas."
-
"Customer tables."
-
"Tables with claim in the name."
Adding more tables to an existing selection
-
"Also add ORDERS and INVOICES tables."
-
"Include AUDIT tables too."
-
"Add tables starting with PROD_ to the previously selected ones."
Best Practices
-
Clearly specify your replication goals and preferences for optimal AI guidance.
-
Review AI-generated mappings and configurations before saving the pipeline.
-
Use incremental intent refinement for complex or evolving requirements.
-
Leverage AI-driven modification to update existing pipelines efficiently.
Results
The Automated Replication Pipeline Designer accelerates pipeline creation, reduces manual configuration, and ensures accurate, consistent replication processes. By leveraging AI guidance, you minimize errors and gain confidence in your data integration workflows.