CX Aperitivo — our next London meet-up is Thursday 17 September. Save your spot.

CX Aperitivo — our next London meet-up is Thursday 17 September. Save your spot.

Welcome to agent-led growth

Welcome to agent-led growth

Build your own in-product onboarding agent that converts new users.

Build your own in-product onboarding agent that converts new users.

Build your own in-product onboarding agent that converts new users.

INSTALL THE SDK · 5 MIN

QUICK CHAT

Experience the agent built into Fin AI

WHY BUILD THIS?

A better first impression becomes long-term growth.

A better first impression becomes long-term growth.

A better first impression becomes long-term growth.

01

Higher trial conversion

Give each new user a clear path to their first real win.

02

Attention span is falling

Due to social media habits users evaluate a platform's value in seconds

03

First impression lasts

You have one shot at a good first impression inside the product before they leave

LEARN MORE

HOW IT WORKS

One MCP connection to your existing chatbot agent.

One MCP connection to your existing chatbot agent.

One MCP connection to your existing chatbot agent.

Your chatbot agent

Already knows your product

Autoplay MCP

one connection

Autoplay onboarding agent

Guides users contextually in the moment

LEARN MORE

WHERE SOFTWARE IS GOING

Software is becoming a hybrid experience between chat and UI.

Software is becoming a hybrid experience between chat and UI.

Software is becoming a hybrid experience between chat and UI.

01

01

Prompt

Prompt

Someone asks for the outcome they want, in their own words.

Someone asks for the outcome they want, in their own words.

02

02

Agent takes action

Agent takes action

The agent does the work in-product — no hand-off, no help center.

The agent does the work in-product — no hand-off, no help center.

03

03

UI refinement and navigation

UI refinement and navigation

The interface assembles around the request and moves them to the right place.

The interface assembles around the request and moves them to the right place.

04

04

Back to prompt

Back to prompt

They ask the next thing, and the loop tightens.

They ask the next thing, and the loop tightens.

HOW IT FITS TOGETHER

We build one unified context layer
powering your onboarding agent.

Your analytics already know what every user does. Your agent already knows what it has done. Autoplay MCP joins the two and hands that context to the chatbot you already run — so it can suggest the one next step that unlocks value, then show the user how, or simply do it for them.

Live user activity
from product analytics
Coming soon
Agent action history
MCP tool history trace
Autoplay MCP
one connection
Your chatbot agent
the one you already run
Suggests the next best step
to get real value out of the product
Show me
visual guidance, in-app
Do it for me
the agent takes the action
See it in action

FAQ

What is an onboarding agent, in plain terms?

How is this different from in-app guidance tools like Pendo and Appcues?

Does this replace our customer support tool?

What does the setup look like — do we need engineering?

What’s the actual impact?

When is it a good time to adopt this?

Why not just continue with product demos and videos?