Skip to Content
Architecture

System Architecture

This page maps out the full request lifecycle in MeetAI — from a user clicking “Start Meeting” to receiving an AI-generated summary in their dashboard.

High-Level Architecture

Request Lifecycle

1. Meeting Creation

2. Live Meeting — Audio & Transcript Flow

3. Post-Meeting Processing

Component Responsibilities

ComponentRoleKey Files
Next.js App RouterSSR, API routes, tRPC serversrc/app/, src/trpc/
tRPCTypesafe API layer for all CRUDsrc/trpc/routers/
Better AuthOAuth + email auth, session mgmtsrc/lib/auth.ts, src/middleware.ts
LiveKit Server SDKRoom creation, token generationsrc/lib/stream-video.ts
LiveKit AgentVoice AI pipeline, tool callingmeetai-agent/src/agent.ts
InngestEvent-driven post-meeting processingsrc/inngest/functions.ts
Drizzle ORMTypesafe DB queries over Neonsrc/db/schema.ts, src/db/index.ts
Polar.shSubscription billing, premium gatingsrc/lib/polar.ts

Data Flow Summary

  1. Write path: Agent → HTTP POST → Next.js API → Drizzle → Neon (transcript lines batched every 3s)
  2. Read path: Dashboard → tRPC → Drizzle → Neon → React Server Components
  3. Real-time path: Agent → LiveKit Data Channel → Browser (live captions, no DB round-trip)
  4. Async path: LiveKit webhook → Inngest event → Durable function steps → Gemini → Neon
Last updated on