A database query — written, then proven by running it

Say in plain words what you need from the database — the scenario writes the query, runs it in a safe test environment and fixes it itself if it fails.

  • The query is written from a plain-language ask
  • Proven by running it, not by trust
  • Errors are fixed automatically
  • Your live database is never touched
Set it upScriptera

How it works

To connect Databricks and Microsoft SQL, you don't need a developer: a ready-made scenario links them in minutes.

  1. Starts when: When chat message received
  2. Then: localVariables
  3. Check: IfOpenAI
  4. If yes: AgentName
  5. Then: getAssistantsList
  6. Check: If
  7. If yes: isAssistantExistsCode
  8. Check: isAssistantExists
  9. If yes: assistant
  10. Then: OpenAIMainBrain
  11. Then: setThreadId
  12. Check: isExecutable
  13. If yes: setOutputByProvider
  14. Check: IfError
  15. If yes: AutoErrorFixing
  16. If yes: IsMaxAutoErrorReached
  17. If yes: collectErrorLoopDetails1
  18. Then: maxAutoErrorLimitReached
  19. If no: Switch
  20. If yes: Execute_AI_result
  21. Then: mergeExecutionsResult
  22. Then: collectErrorLoopDetails
  23. Then: executedSQLQuery
  24. Then: GenerateErrorPrompt
  25. Check: isOpenAI
  26. If no: OpenRouterAgent
  27. If no: Execute a SQL query
  28. If no: askUserHowToHandleError
  29. If no: wordsForUser1
  30. If no: createOpenAiAssistant
  31. If no: issueOnOpenAiSide
  32. Then: Simple Memory
  33. Then: OpenRouter Chat Model

You can launch this Databricks + Microsoft SQL integration in Scriptera: describe the task in plain words — the scenario is built, launched and monitored for you.

Similar scenarios