Let AI schedule meetings for you

One tool. One API. Fully automated scheduling.

No slot picking. No back-and-forth. Just results.

Try AI booking nowView MCP manifest

Try MCP in action

Idle

Bookora lets AI agents turn natural language into confirmed bookings — in one step.

API base: https://api.bookora.work
User
Book a meeting tomorrow at 3pm
AI (Claude / GPT)
Waiting...
MCP Tool
POST /mcp/v1/execute
Result
—
Live flow
Input
“Schedule a meeting next week afternoon”
AI Action
create_booking()
Result
✅ Tue, 2:00 PM booked
Try it with one request
POST /mcp/v1/execute

{
  "tool": "create_booking",
  "input": {
    "intent": "book_meeting",
    "time_preference": "next week afternoon",
    "duration": 30
  }
}
Send this request and get a confirmed booking instantly.

What Bookora MCP adds

  • AI-native scheduling: agents express intent, Bookora executes decisions
  • Automatic slot selection: no manual slot picking
  • Learning system: stores decision metadata (propensity, weights, features)
  • Strategy layer: controls optimization behavior (feature-flagged)
Works with
  • OpenAI (GPT)
  • Anthropic (Claude)
  • Any MCP-compatible agent
Compatible with modern AI tool calling systems

create_booking (AI-first)

This is not a typical booking API.

  • Understands natural language intent
  • Selects the best time automatically
  • Handles conflicts and constraints
  • Creates the booking in one step
Key message
You don’t pick a time.
Bookora does.
Bookora is not a scheduling API. It’s an AI-native scheduling engine that turns intent into a confirmed booking.

Why not traditional scheduling tools?

Traditional tools (e.g. Calendly)
  • Require manual slot selection
  • Expose low-level APIs
  • Push complexity to developers
Bookora
  • AI selects the best time
  • One tool handles everything
  • Built for GPT, Claude, and AI agents

What you can build

AI assistants
Schedule meetings directly from chat.
Booking chatbots
Handle booking automatically end-to-end.
Internal tools
Smart scheduling in ops, sales, and support flows.

How it connects

MCP lets AI call tools instead of generating text.

User input
  ↓
AI (GPT / Claude)
  ↓
Tool call (MCP)
  ↓
POST /mcp/v1/execute
  ↓
Bookora scheduling engine (/ai/schedule)
  ↓
Booking confirmed

Each booking improves future decisions through the learning loop.

Try AI booking with one request

Endpoint
POST /mcp/v1/execute
{
  "tool": "create_booking",
  "input": {
    "intent": "book_meeting",
    "time_preference": "next week afternoon",
    "duration": 30
  }
}

Capability Layers

Execution layer

  • MCP tool execution (/mcp/v1/execute)
  • Booking mutations (create / cancel)

Decision layer

  • Scheduling engine selects slots automatically

Intelligence layer

  • AI decision system (LinUCB)
  • Memory system
  • Strategy layer
Other tools (lower-level)
get_availability — returns available time slots
cancel_booking — cancels an existing booking

AI Scheduling Layer

  • Contextual bandit (LinUCB)
  • Exploration vs exploitation (alpha)
  • Real-time optimization (feature-flagged)
  • Multi-tenant model isolation (feature-flagged)

Memory & Learning System

  • Stores weights and metadata in bookings
  • Logs behavior and decision context
  • Supports IPS (propensity-based evaluation)
  • Uses best-weights.json as fallback

Strategy Layer

  • Optimizes for revenue / utilization / satisfaction
  • Per-org and per-booking-type configuration
  • Controls exploration, time windows, preferences

MCP manifest

Endpoint
GET /mcp/v1/manifest
Returns tool definitions, a system prompt, few-shot examples, and constraints your agent should follow.

Architecture

MCP interface

  • /mcp/v1/manifest
  • /mcp/v1/execute

Scheduling engine

  • Availability → candidate slots → selection (/ai/schedule)

AI decision layer

  • LinUCB scoring + exploration

Memory layer

  • Booking metadata + learning signals

Ready to ship AI scheduling?

Connect your agent once and let Bookora handle the rest.

Schedule your first meeting with AITry with one request
Looking for traditional docs? See API docs