Welcome to value-led growth!
Initialize the SDK in your frontend, identify your users, and Autoplay starts streaming every interaction — clicks, views, flows — to your copilot in real time.
Read the full Quickstart →import posthog from 'posthog-js'
posthog.init('YOUR_AUTOPLAY_API_KEY', {
api_host: 'https://us.i.posthog.com',
person_profiles: 'identified_only',
session_idle_timeout_seconds: 120,
loaded: (posthog) => {
posthog.identify(posthog.get_distinct_id(), {
product_id: 'YOUR_AUTOPLAY_PRODUCT_ID',
});
},
})Stream real-time user events into your AI stack with a single client.
Reliable event buffering so nothing gets lost between app and copilot.
Live summaries of the user’s session to ground every reply.
Rich, structured context so your agent knows history, workflows, and intent.
While they’re hesitating on pricing, stuck on onboarding, or about to abandon a flow — your copilot already knows. And acts.
Proactive copilot support means your AI agent reaches out to users before they ask for help. Autoplay streams real-time behavior — clicks, hesitation, repeated errors, abandoned flows — into your copilot, so it can intervene the moment a user gets stuck instead of waiting for a ticket.
Standard chatbots are reactive — they only respond after the user types a message. Autoplay gives your copilot real-time situational awareness of what every user is doing, so it can trigger help proactively, personalize responses with full session context, and resolve issues before they escalate.
Autoplay integrates natively with Intercom, Zendesk, HubSpot Chat, Help Scout, Dify, and Botpress. The AgentContext API also makes it easy to plug into any custom LLM agent or copilot.
Most teams ship a working proactive copilot in under an hour using the Autoplay quickstart. Drop in the AsyncConnectorClient, connect your existing support tool, and your copilot starts receiving real-time user context immediately.