Converting ETL

Converting ETL means moving legacy data integration jobs to Snowflake by separating transformation (how data moves between sources and targets) from orchestration (when jobs run, dependencies, and variables). On Snowflake, transformation typically lands in dbt projects on Snowflake and orchestration maps to native Snowflake constructs such as tasks and stored procedures. The Snowflake AIM Agent for Data Warehouses drives this conversion for SSIS and Informatica PowerCenter workloads.


ETL conversion in the Snowflake AIM Agent for Data Warehouses

The agent treats ETL as a first-class capability alongside SQL conversion, assessment, and data migration:

CapabilityWhat it means
Convert ETL pipelinesTranslate SSIS packages and Informatica workflows toward dbt using deterministic conversion, with optional AI-assisted remediation for remaining gaps
Assess workloadsThe interactive HTML report includes SSIS / Informatica ETL analysis (package classification, control- and data-flow mapping, effort estimates) alongside waves, exclusions, and dynamic SQL
Output layoutReplatform output lands under Output/ETL/ with shared etl_configuration/ plus per-package folders that combine orchestration SQL (tasks or procedures) with per-data-flow dbt projects (staging, intermediate, marts)

The conceptual mapping from source ETL constructs to Snowflake artifacts looks like this:

Source conceptTypical output role
SSIS Data Flow TaskStandalone dbt project (transformation)
SSIS Control FlowTASK or procedure (orchestration)
SSIS variablescontrol_variables-style infrastructure plus dbt variables
Informatica mapping / workflowdbt models plus orchestration tasks

Example prompts (from the skill topic):

"Assess my SSIS packages"
"Convert my SSIS packages"

Platform availability for SSIS to dbt, Informatica to dbt, and the corresponding AI conversion paths appears in the Supported source systems table in the Snowflake AIM Agent for Data Warehouses (deterministic and AI-assisted conversion are available across SQL Server, Redshift, Teradata, Oracle, and other source systems).