Messenger assistant: questions, orders and tracking

How it works

To connect Postgres PGVector Store and exolve, you don't need a developer: a ready-made scenario links them in minutes.

1. Starts when: WhatsApp Incoming Message Trigger
2. Then: Extract WhatsApp Message Details
3. Then: AI Intent Classification
4. Then: Parse Intent Classification Response
5. Check: Intent-Based Request Router
6. If yes: WhatsApp FAQ AI Agent
7. Then: Send FAQ WhatsApp Reply
8. If no: AI Order Information Extractor
9. Then: Parse Order Details JSON
10. Then: Fetch Product Catalog
11. Then: Prepare Product Catalog Context
12. Then: AI Semantic Product Matcher
13. Then: Parse & Validate Product Match
14. Check: Product Match Validation Check
15. If yes: Inventory Availability Check
16. If yes: Create New Order Record
17. Then: Update Product Inventory
18. Then: Send Order Confirmation Message
19. If no: Send Product Error / Stock Failure Message
20. If no: AI Order ID Extractor
21. Then: Parse Order Tracking Request
22. Then: Fetch Order Status
23. Then: Send Order Status Reply
24. If no: Notify Human Support Team
25. Then: Send Human Support Acknowledgement
26. Then: FAQ Response Language Model
27. Then: Conversation Memory Manager
28. Then: Knowledge Base Retrieval Tool
29. Then: Pinecone FAQ Vector Store
30. Then: FAQ Embedding Generator
31. Then: Knowledge Retrieval LLM
32. Starts when: Knowledge Base File Change Trigger
33. Then: Download Knowledge Base File
34. Then: Store Knowledge Embeddings In Pinecone
35. Then: Split Knowledge Base Into Chunks
36. Then: Load Knowledge Base Document
37. Then: Generate Knowledge Embeddings

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