Support chat: AI answers, hard cases go to a live agent

The scenario answers support chats from the knowledge base itself, and escalates complex conversations to a live agent while alerting the team.

  • AI closes simple questions from the knowledge base
  • Complex cases escalate to a live agent
  • The team is alerted with an assigned agent
  • People handle only what needs people
Postgres Chat MemoryGmail ToolWebhookPinecone Vector StoreEmbeddings Google Gemini
Connect Postgres PGVector Store and runScriptera

How it works

To get SendGrid notifications in Postgres PGVector Store, you don't need a developer: a ready-made scenario watches events and sends messages for you.

  1. Starts when: Incoming Chat Webhook
  2. Then: Prepare Field Edits
  3. Check: Route by Chat Message Type
  4. If yes: Check Live Agent Assignment
  5. If no: Fetch Live Agent Data
  6. Then: Generate Assignee Agent ID
  7. Then: AI Customer Support Agent
  8. Then: Split AI Messages
  9. Then: Store Chat Memory in Postgres
  10. Then: Send Internal Team Email
  11. Then: Escalate Chat via HTTP
  12. Then: Use Groq Chat Model
  13. Then: Utilize Pinecone Vector Store
  14. Then: Generate Google Gemini Embeddings
  15. Then: Implement Google Gemini Chat Model
  16. Then: Access Knowledge Base

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

Similar scenarios