A sales assistant across every messenger

The assistant answers customers in messengers and social networks from your knowledge base, logs them to Google Cloud Realtime Database and books meetings in Google Calendar.

  • Answers in every channel around the clock
  • Replies from your knowledge base
  • Logs customers to a table
  • Books meetings in the calendar
Postgres Chat MemoryPostgres PGVector StoreWhatsApp Business CloudWebhook
Connect Pgvector and launchScriptera

How it works

To move data from Pgvector to Telegram automatically, use a ready-made scenario — no manual exports.

  1. Starts when: When chat message received
  2. Then: Set Message6
  3. Then: Update Agent Details
  4. Then: Sales Agent
  5. Check: Switch1
  6. If yes: Send message (Whatsapp)
  7. If no: Send message (facebook)
  8. If no: Send message (instagram)
  9. If no: Send message (embedded)
  10. If no: Send message (telegram)
  11. Then: Postgres Chat Memory
  12. Then: Consultation Agent
  13. Then: CRM Agent
  14. Then: Vector Database
  15. Then: Company Database
  16. Then: Reply To User1
  17. Starts when: Telegram Trigger
  18. Check: Route Message
  19. If yes: Set Message
  20. If no: Get Audio1
  21. Then: Transcribe1
  22. Then: Set Message1
  23. If no: Reply to user
  24. Starts when: Facebook Trigger
  25. Starts when: Respond to Webhook - facebook post
  26. Check: If not echo
  27. If yes: Set Message5
  28. If yes: Facebook Typing Effect
  29. Starts when: Respond to Webhook - facebook get
  30. Starts when: Instagram Trigger
  31. Starts when: Respond to Webhook - instagram post
  32. Check: If not echo1
  33. If yes: Set Message4
  34. Starts when: Respond to Webhook - instagram get
  35. Starts when: WhatsApp Trigger
  36. Then: Get Lead Info
  37. Check: Route Messages
  38. If yes: Get Audio
  39. Then: Download Audio
  40. Then: Transcribe
  41. Then: Set Message2
  42. If no: Set Message3
  43. Starts when: When clicking ‘Execute workflow’
  44. Then: Download file
  45. Then: PGVector
  46. Then: Default Data Loader
  47. Then: Gemini
  48. Then: Response
  49. Then: Try Again
  50. Starts when: When Executed by Another Workflow
  51. Then: CRM Assistant
  52. Then: get_schema
  53. Then: list_records
  54. Then: Window Buffer Memory
  55. Then: create_opportunity
  56. Then: update_opportunity
  57. Then: create_contact
  58. Then: update_contact
  59. Then: Success
  60. Then: Calendar Agent
  61. Then: Try Again1
  62. Then: Create Event with Attendee
  63. Then: Get Events
  64. Then: Delete Event
  65. Then: Update Event
  66. Then: OpenAI Chat Model
  67. Then: OpenAI Chat Model1
  68. Then: OpenAI Chat Model2
  69. Then: OpenAI Chat Model3
  70. Then: Embeddings OpenAI

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

Similar scenarios