Skip to main content
This section provides a technical reference for how OpenBat is structured, useful for developers integrating with the platform or understanding its capabilities.

​
Tech stack

LayerTechnology
FrontendNext.js (App Router), shadcn/ui + Tailwind CSS
ChartsCharting library (sentiment trends, donut charts, activity heatmaps)
BackendNext.js Route Handlers / Server Actions
AI pipelineLLM-based analysis (prompt-driven, configurable per chatbot)
Semantic searchVector embeddings + similarity search
AuthenticationEmail/password + social OAuth (Apple, Google, Meta)

​
URL structure

/                                          → Public marketing site
/auth/*                                    → Authentication (login, sign-up, forgot-password)
/platform                                  → Organization-scoped chatbot list
/platform/members                          → Organization members
/platform/settings                         → Organization settings
/platform/{chatbotId}                      → Chatbot analytics dashboard
/platform/{chatbotId}/workflows            → Workflow automation
/platform/{chatbotId}/conversations        → Conversation browser
/platform/{chatbotId}/conversations/{id}   → Conversation detail
/platform/{chatbotId}/deep-search          → Semantic search
/platform/{chatbotId}/organizations        → Client organizations
/platform/{chatbotId}/users                → Client users
/platform/{chatbotId}/analysis-config      → Analysis configuration
/platform/{chatbotId}/settings             → Chatbot settings
The URL structure reflects a two-level hierarchy: organization (at /platform) and chatbot (at /platform/{chatbotId}). All chatbot sub-resources are scoped to the chatbot ID.

​
Authentication model

LevelScopeRolesURL
OrganizationAll chatbots in the orgOwner, Admin, Member/platform/members
ChatbotSingle chatbotMember (addable)/platform/{chatbotId}/settings > Members
Authentication methods:
  • Email/password — standard registration and login with password requirements (8+ characters)
  • Social OAuth — Apple, Google, and Meta via OAuth redirect flows
  • Password reset — self-service via email
The sidebar organization switcher indicates the authenticated session carries organization context, allowing seamless switching between organizations.

​
API surface

​
External API (SDK to OpenBat)

  • Conversation ingestion endpoint (authenticated via API key)
  • Message data with user/assistant roles
  • Custom metadata fields
  • Session context (page URL, referrer, device, country)

​
Internal API (frontend to backend)

  • CRUD operations for organizations, chatbots, members, workflows, webhooks
  • Analytics queries for dashboard KPIs, sentiment trends, resolution outcomes
  • Search: keyword filtering and semantic deep search
  • Configuration: analysis flags, prompts, translation settings
  • Data import and export

​
Outbound integrations

  • Webhook delivery to Discord, Slack, and custom HTTP endpoints
  • Social OAuth provider communication
  • AI/LLM provider calls for the analysis pipeline

​
Next steps

Data model

Entity relationships and multi-tenancy design.

Analysis pipeline

How conversations are analyzed end-to-end.