Unified SQL processing engine: transpilation, optimization, batch translation, and REST API — 500+ function mappings, stored procedure migration, and custom parsers built for each dialect.
15+ source dialects → 7+ cloud targets
SQLForge goes beyond string replacement: CTE and join optimization, window and aggregation tuning, Teradata-style calculated fields, subquery-to-CTE conversion, query splitting, index suggestions, security scanning, and cost profiling — all callable from the same service surface as translation.
Full-statement transpilation with per-dialect function mapping. Batch endpoints for folders of SQL, plus GET-based translation for quick tests.
/translate-sql and /translate-sql-batch/map-function for single-function parity checks/validate-sql for syntax checks by dialectRewrite queries for clearer plans: CTE folding, join reordering, aggregation and window optimizations, parallel hints where applicable.
/optimize-sql with strategy optionsBreak very long SQL into maintainable pieces, combine scripts as CTE chains, and convert correlated subqueries to readable CTEs.
/sqlbreaker, /split-query/convert-subqueries-to-cte/combine-queriesDetect tables and columns from SQL text, surface calculated fields, and emit DDL-oriented hints for documentation and migration packs.
/detect-schema/detect-calculated-fieldsStatic analysis for injection patterns and risky constructs; maintenance hints and query profiling to estimate cost and bottlenecks.
Integrate legacy SAS and macro-heavy pipelines: macro transforms and SQL extraction paths that feed the same translation stack.
/sqlmacros for SAS macro processingRun SQLForge as a single process (for example uvicorn) and integrate from IDEs, pipelines, or internal portals. Same routes support interactive Swagger UI and machine clients.
/translate-sql
Translate full SQL between dialects with function mapping applied.
/translate-sql-batch
Multiple statements or files in one request for bulk migration jobs.
/supported-dialects
List source and target dialects available for the running build.
/validate-sql
Syntax validation for a chosen dialect before execution on target.
/optimize-sql
Optimization passes: CTEs, joins, aggregations, and more.
/health
Liveness for load balancers and Kubernetes probes.
Full endpoint list includes encode/decode helpers, dialect conversion, index suggestions, SQL breaking, and unified / root discovery — see OpenAPI when deployed.
MigryX builds a dedicated parser per dialect — not a lowest-common-denominator grammar. Every dialect-specific construct is parsed, classified, and migrated with full fidelity.
MigryX builds a dedicated parser for each source dialect. This eliminates the parse failures common in generic ANTLR-based tools and enables accurate AST construction for complex stored procedures, CTEs, and dialect-specific syntax.
Every layer of SQL migration is handled — from DDL and DML through stored procedures, views, UDFs, and schema dependency ordering.
Dialect-specific scalar and table-valued functions are resolved to target equivalents. Where no native equivalent exists, MigryX generates a Python / JavaScript UDF scaffold for the target platform.
Every migrated SQL object carries forward column-level lineage metadata — source table, source column, transformation expression — surfaced as STTM / data catalog exports.
MigryX maintains a curated, tested mapping library covering date/time functions, string functions, NULL coalescing, aggregation, type conversion, and dialect-specific syntax rewrites.
| Source Dialect | Source Function / Syntax | Target Dialect | Target Equivalent | Category |
|---|---|---|---|---|
| Oracle | NVL(col, 0) | Snowflake / BigQuery / Databricks | COALESCE(col, 0) | NULL handling |
| Oracle | DECODE(x, a, b, c) | All targets | CASE WHEN x=a THEN b ELSE c END | Conditional |
| Oracle | ROWNUM <= N | Snowflake / BigQuery | ROW_NUMBER() OVER(...) <= N / LIMIT N | Row limiting |
| Oracle | SYSDATE | All targets | CURRENT_TIMESTAMP | Date/time |
| Oracle | ADD_MONTHS(dt, n) | Snowflake | DATEADD(month, n, dt) | Date arithmetic |
| Oracle | TRUNC(dt, 'MONTH') | BigQuery | DATE_TRUNC(dt, MONTH) | Date truncation |
| Oracle | CONNECT BY PRIOR | Snowflake / BigQuery | WITH RECURSIVE ... UNION ALL | Hierarchical |
| T-SQL | ISNULL(col, 0) | All targets | COALESCE(col, 0) | NULL handling |
| T-SQL | GETDATE() | All targets | CURRENT_TIMESTAMP | Date/time |
| T-SQL | TOP N | Snowflake / BigQuery / Databricks | LIMIT N | Row limiting |
| T-SQL | CONVERT(VARCHAR, dt, 101) | Snowflake | TO_VARCHAR(dt, 'MM/DD/YYYY') | Type conversion |
| T-SQL | DATEDIFF(day, d1, d2) | BigQuery | DATE_DIFF(d2, d1, DAY) | Date arithmetic |
| Teradata | QUALIFY ROW_NUMBER()=1 | Snowflake (native) | QUALIFY ROW_NUMBER()=1 | Window filter |
| Teradata | QUALIFY ROW_NUMBER()=1 | BigQuery / Databricks | Subquery with CTE wrapping | Window filter |
| DB2 | FETCH FIRST 10 ROWS ONLY | All targets | LIMIT 10 | Row limiting |
| DB2 | CURRENT DATE | All targets | CURRENT_DATE | Date/time |
| Hive | LATERAL VIEW EXPLODE(arr) | Snowflake | LATERAL FLATTEN(arr) | Array expansion |
| Hive | LATERAL VIEW EXPLODE(arr) | BigQuery | UNNEST(arr) | Array expansion |
| Hive | COLLECT_SET(col) | Snowflake | ARRAY_AGG(DISTINCT col) | Aggregation |
| Netezza | AGE_IN_YEARS(dt) | Snowflake | DATEDIFF('year', dt, CURRENT_DATE) | Date arithmetic |
| Vertica | APPROXIMATE_COUNT_DISTINCT(col) | Snowflake | APPROX_COUNT_DISTINCT(col) | Approximation |
| PostgreSQL | col::INT (cast shorthand) | Snowflake / BigQuery | CAST(col AS INT) | Type conversion |
Showing 22 of 500+ mappings. Full mapping library available in the MigryX assessment report.
MigryX follows a structured migration methodology that covers discovery through validation — with every decision auditable and every migration fully documented.
Inventory all SQL artifacts and build a complete dependency graph before a single line is rewritten.
Apply the mapping library to rewrite every SQL object to target-dialect equivalent syntax.
Confirm semantic equivalence — not just syntactic validity — before migration is signed off.
MigryX runs wherever your data lives. Enterprise clients with restricted environments can deploy fully air-gapped with no outbound internet access required.
Full MigryX platform deployed inside your network perimeter. No source code, SQL, or metadata leaves your environment. Suitable for regulated industries (FSI, Healthcare, Government).
Browser-based MigryX platform with SOC 2 controls. Tenant-isolated processing. Connect your source environment via secure agent or upload SQL artifact packages directly.
We analyze your SQL inventory, classify dialect patterns, count object types, and produce a migration readiness report — before any contract is signed.
MigryX is the only platform with dedicated per-dialect parsers built by SQL migration engineers — not generic grammar tools. Our clients migrate 95%+ of SQL objects automatically.
Book a 30-minute SQL migration discovery call with our engineering team.
Schedule on Calendly