Paste CSV directly or upload a CSV file. The planner inspects each column, infers likely SQL types, suggests a primary key, calls out possible foreign-key or index hints, and then produces database-specific DDL.
How to use it:
- CSV Input or CSV File: provide one dataset with a header row
- Table Name: choose the target table name that should appear in the generated SQL
- Target Database: switch the output dialect between PostgreSQL, MySQL, SQLite, and SQL Server
- Existing Schema: optional current CREATE TABLE or DDL text used to generate ALTER ADD statements for newly inferred columns
- Schema Overrides (JSON): optional JSON map for manual fixes, for example {"price":"DECIMAL(14,4)","customer_id":{"type":"BIGINT","index":true}}
- Generate ALTER Plan: when enabled, the report also lists missing-column ALTER statements when an Existing Schema is supplied
What gets inferred:
- Text vs integer vs decimal vs boolean vs date vs datetime columns
- Nullable vs not-null columns
- Primary-key candidates such as id, *_id, code, or uuid when values are unique
- Foreign-key hints for columns like userid or productid
- Index suggestions for high-cardinality lookup columns
Notes:
- The tool works from one CSV at a time, so foreign keys are advisory hints rather than guaranteed references
- Overrides let you tighten types or force index/reference suggestions before handing the SQL to an engineer or DBA