Transform Your App Into Intuitive AI Experience
Open source AI agents that operate your React application through natural conversation. Users achieve results faster, reducing the learning curve and support needs.
From Chat to Action
InAppAI represents a paradigm shift: AI that doesn't just answer questions, but takes action through your APIs and tools.
Traditional Chatbots
User asks question
AI provides text answer
User reads response
User manually takes action
InAppAI Agents
User states intent
Agent calls React tool
Tool updates app state
UI reflects changes
Three Core Concepts
InAppAI is built on three fundamental concepts that work together to enable agents to operate your UI.
Agents
AI that can perceive your application state, reason about user intent, and operate your UI by calling tools you provide.
Context
How agents understand your app—current state, user data, business rules, conversation history. Context enables intelligent decisions.
Tools
Functions you expose to agents—create, update, delete, filter, search. Agents call your tools to manipulate application state and update the UI.
What You Can Build
Imagine agents that understand your app and take action on behalf of your users
Payment Processing Dashboard
User Intent
"Show me a complete overview of last week's issues"
Agent Actions
- 1 Agent sees context: Dashboard needs failed payments refunds and chargebacks
- 2 Calls 3 tools in parallel: getFailedPayments() getRefunds() getChargebacks()
- 3 All data fetched simultaneously in 1.2s instead of 3.6s sequential
- 4 Dashboard updates with all three data sets
Result
Complete overview: 23 failed payments, 12 refunds, 5 chargebacks
Sales CRM
User Intent
"Schedule follow-up with John Smith for next Tuesday"
Agent Actions
- 1 Agent sees context: Contact 'John Smith' ID #2847 in customer list
- 2 Calls 3 tools in parallel: createCalendarEvent() sendEmailNotification() createFollowUpTask()
- 3 Event notification and task created simultaneously
- 4 All systems updated and contact informed
Result
Meeting scheduled, email sent, task created - all in parallel
E-commerce Admin
User Intent
"Mark all winter jackets as clearance with 30% off"
Agent Actions
- 1 Agent sees context: 47 products tagged 'winter-jackets'
- 2 Calls bulkUpdateProducts tool with tag='winter-jackets' clearance=true discount=0.30
- 3 All matching products updated and marked live
Result
47 winter jackets marked clearance with 30% discount
Built for Developers
Complete toolkit with React components, backend APIs, and multi-provider AI support
TypeScript-first API with full type safety
React components for agent UI
Backend API for agent reasoning and tool orchestration
Multi-provider AI support (OpenAI Anthropic Google)
Knowledge base integration for domain context
Usage tracking and permissions enforcement
Open source React components
Production-ready and battle-tested
Ready to Get Started?
Choose your path to building agent-operated UIs
Read the Docs
Learn how agents, tools, and context work together to enable intelligent UI operations
View Documentation