MigryX parses every .dtsx package and .ispac project archive — data flow components, control flow tasks, SSIS expressions, connection managers, and C#/VB.NET script tasks — then generates production-ready code for modern cloud platforms.
MigryX reads SSIS packages directly from the file system or the SSISDB catalog — extracting every component, expression, variable, connection, and script task for complete automated conversion.
MigryX understands the SSIS data flow graph at the component level, preserving transformation semantics when generating pandas, PySpark, or Snowflake SQL output.
OLE DB Source, ADO NET Source, Flat File Source, Excel Source, XML Source, Raw File Source, ODBC Source. Connection strings extracted and converted to cloud-native connectors.
Derived Column, Conditional Split, Lookup, Merge Join, Sort, Aggregate, Data Conversion, Character Map, Copy Column, Multicast, Union All, Pivot, Unpivot, Slowly Changing Dimension (SCD Type 1/2/3).
OLE DB Destination, ADO NET Destination, Flat File Destination, Excel Destination, Raw File Destination, SQL Server Destination, ODBC Destination. Bulk load patterns preserved.
Full SSIS expression language parsing — string functions, date functions, type casts, conditional operators, and variable/parameter references. Resolved to Python or Jinja2 at conversion time.
Slowly Changing Dimension wizard output fully parsed and rewritten. Type 1 (overwrite), Type 2 (history rows), and Type 3 (alternate column) converted to MERGE statements or dbt snapshot models.
Lookup Transformation (full-cache, partial-cache, no-cache modes), Cache Transform, Fuzzy Lookup, Term Lookup — converted to JOIN operations or in-memory dictionary lookups in Python.
Every SSIS control flow element — precedence constraints, loops, containers, and event handlers — is modeled and converted to equivalent orchestration patterns.
SQL statements, stored procedure calls, and result sets mapped to Python database operators or ADF pipeline activities. Connection strings resolved to target connectors.
File enumerator, ADO object enumerator, and item collection loops. Converted to Airflow dynamic task mapping or Python iteration patterns with parameterized sub-tasks.
Counter-based iterations with InitExpression, EvalExpression, and AssignExpression. Converted to Python while/for loops or Airflow sensor + trigger patterns.
Task grouping, rollback boundaries, and transaction scope. Mapped to Airflow TaskGroup or Python context managers preserving execution order and error propagation.
Inline C# and VB.NET scripts decompiled and analyzed for external dependencies. Converted to Python with AI-assisted translation and a manual review checklist for complex logic.
Child package invocations and package-level parameter passing. Converted to sub-DAG calls in Airflow or modular Python function calls with parameter injection.
OnError, OnWarning, OnTaskFailed, OnPostExecute event handlers converted to Airflow callbacks, Python exception handlers, or alerting logic in Prefect.
SQL Server Agent job steps invoking SSIS packages via dtexec or SSISDB catalog execution. Schedules, parameters, and invocation patterns converted to Airflow or Prefect schedules.
MigryX scans your .dtsx files, .ispac archives, and SSISDB catalog. Every package receives a complexity score and a migration effort estimate before conversion begins.
The MigryX parser engine converts each package — data flow, control flow, expressions, and connection configuration — to idiomatic code for your chosen target platform.
Every converted pipeline is accompanied by a row-level validation suite. MigryX runs converted code in parallel with the original to confirm output parity before handover.
Key transformation mappings MigryX applies automatically. Every mapping is documented in the lineage report.
| SSIS Component | Python / pandas | Airflow | Azure Data Factory |
|---|---|---|---|
| OLE DB Source | SQLAlchemy + pd.read_sql() | SQLExecuteQueryOperator | Linked Service + Dataset |
| Derived Column | df.assign() / df[col] = expr | Python operator inline | Data Flow derived column activity |
| Conditional Split | df[df[condition]], df.query() | BranchPythonOperator | Conditional Activity / If Condition |
| Lookup | df.merge(lookup_df, how='left') | Python operator with join | Lookup activity |
| Merge Join | pd.merge(df1, df2, on=key) | Python operator | Join in Mapping Data Flow |
| SCD Wizard (Type 2) | MERGE + effective_date tracking | Python + upsert logic | Slowly Changing Dimension activity |
| ForEach Loop | for item in items: ... | Dynamic task mapping @expand() | ForEach activity |
| Execute SQL Task | engine.execute(sql) | SQLExecuteQueryOperator | Stored Procedure activity |
MigryX generates idiomatic, runnable code — not a generic approximation. Each target platform has its own ruleset tuned for best practices and performance.
TaskFlow API DAGs, dynamic task mapping from ForEach loops, schedules from SQL Agent jobs, sensor patterns for file watchers.
Full ETL scripts with SQLAlchemy connections, pandas DataFrames, type annotations, and logging. Runnable standalone or as Airflow operators.
Distributed DataFrame transformations on Databricks, EMR, or Dataproc — partitioned writes, native Spark SQL, Delta Lake integration.
Prefect flows and tasks preserving control flow logic, with retry policies, logging, and parameterized runs from SSIS package configurations.
ADF pipelines, datasets, linked services, and Mapping Data Flows generated with full parameter mapping from SSIS project-level parameters.
Snowflake SQL with MERGE for SCD patterns, COPY INTO for bulk loads, and Snowpark Python for complex transformations that require procedural logic.
Delta Lake notebooks, Unity Catalog integration, and Databricks Jobs API scheduling from SSIS execution patterns and SQL Agent job definitions.
AWS Glue PySpark ETL jobs with DynamicFrames, Glue Data Catalog integration, and Glue Workflows for orchestration of multi-package pipelines.
MigryX runs entirely inside your environment. No .dtsx files, credentials, or intermediate artifacts leave your network at any point.
Windows Server or Linux VM in your data center. Reads .dtsx files from the file system or SSISDB. No internet required during conversion.
Fully disconnected operation for regulated industries. Self-contained installation package with no external runtime dependencies.
Azure Private VMs, AWS VPC, or GCP private network. Containerized deployment with Docker or Kubernetes for enterprise platforms.
SOC 2, HIPAA, and FedRAMP-compatible deployment patterns. Full audit log of every file parsed, converted, and validated.
Run MigryX against your actual SSIS packages and see the converted output before committing to a full migration program.
Tell us about your SSIS environment — package count, SQL Server version, target platform, and timeline — and we'll scope a pilot that proves value on your actual workloads.
