Cal.com Agent v2 Demo

Cal.com AI Agents

API v2

This demo mirrors the new Cal.com AI agent guidance: use slots, bookings, reschedule, and cancel workflows on v2 endpoints with explicit versioning. It is built as an agent-style skill layer that can be attached to your main chatbot while preserving normal RAG and intent behavior.

Live Slot + Booking Demo

Demo is live. Complete human verification before scheduling actions.

Human verification (required for create, reschedule, cancel)

Scheduling actions are blocked until verification is complete.

Reschedule / Cancel

Use a booking UID. Reschedule and cancel require Cal.com API key permissions.

Prefer direct booking? Use cal.com/virgentai/chat.

Demo Output

Run an action to see live output.

How It Works

1) Detect scheduling intent from natural language.

2) Fetch live availability from `GET /v2/slots`.

3) Create booking with `POST /v2/bookings`.

4) Handle updates via reschedule/cancel endpoints.

Value Props

• Faster scheduling directly from chat.

• Less drop-off versus sending users away too early.

• Better qualification context attached to booking.

• Deterministic skill path for predictable outcomes.

Source Docs

Official Cal.com references used for this demo.

AI Agents guide Slots API Create booking API

Architecture

User Chat
   │
   ▼
Intent Router (main site agent)
   │
   ├── non-scheduling ──► standard RAG + model routing
   │
   └── scheduling intent ─► Cal.com Scheduling Skill
                            │
                            ├── GET /v2/slots
                            ├── POST /v2/bookings
                            ├── POST /v2/bookings/{uid}/reschedule
                            └── POST /v2/bookings/{uid}/cancel

Want this in your production chatbot?

We can wire this Cal.com scheduling skill into your existing assistant without breaking core RAG and intent logic.

Book a build session

Back to all agents