Plain English โ†’ Any Automation
MCP Compatible

AUTOMATE FASTER
THAN YOU CAN THINK

The only automation platform that solves any request โ€” even ones it's never seen before. Powered by AI that generates code on the fly, connects to your Google workspace, and learns with every automation.

118+
Executions Run
30+
Built-in Tools
4
AI Providers
๐Ÿ‡ฎ๐Ÿ‡ณ
Built for Bharat
SmartFlow

Overview

System OperationalMCP: 17 tools live
30+
Registered Tools
17
MCP-Exposed Tools
99.9%
340ms
Avg Execution Time

Recent Workflows

Daily Stock Analysis

fetch_stock_price โ†’ send_email

Running
MCP: Weather Monitor

fetch_weather โ†’ condition โ†’ send_sms

MCP Active

Everything you need to scale

30+ tools, MCP protocol, self-healing AI, and production-grade architecture.

Natural Language AI

Describe your intent in plain English. Google Gemini translates it into executable workflows with auto-retry.

MCP Server

Full Model Context Protocol support. 17 tools discoverable by Claude Desktop, Cursor, and any MCP client.

30+ Tool Registry

Centralized registry with JSON Schema definitions. 7 categories: data, scraping, transform, notification, and more.

Self-Healing Engine

Workflow validator catches errors before execution. AI auto-fixes malformed steps and retries intelligently.

Smart Scheduling

Run cron jobs, set intervals, or trigger from webhooks. Full execution history in Firebase.

Unlimited Capabilities

No fixed templates. If a tool doesn't exist, the AI generates Python code on the fly to solve any request.

Model Context Protocol

MCP Server Built In

Any MCP client can discover and execute all 17 tools via standard protocol. Zero integration code needed.

MCP Request
// Discover all tools
POST /mcp
{
"method": "tools/list"
}
// Execute a tool
POST /mcp
{
"method": "tools/call",
"params": {
"name": "fetch_weather",
"arguments": { "city": "Mumbai" }
}
}

Streamable HTTP Transport

Standard MCP protocol over HTTP POST. Stateless mode โ€” optimized for cloud deployment.

17 Tools with JSON Schema

Every tool has full parameter schemas. Clients auto-discover capabilities โ€” no manual config.

Context Memory & Logging

Each MCP tool call gets execution ID, state tracking, and Firebase persistence.

25s Timeout Protection

Render-compatible request timeout with graceful error handling.

Production-Grade Architecture

Three-layer system: AI engine, workflow executor, and MCP server โ€” all working together.

AI Engine

  • Google Gemini + OpenRouter
  • NL โ†’ JSON workflows
  • Auto-retry with fallback
  • Self-healing validation
  • Context-aware clarification

Workflow Engine

  • 30+ tools in 7 categories
  • Sequential step execution
  • Context Memory passing
  • Firebase state logging
  • Cron scheduling

MCP Server

  • Streamable HTTP transport
  • 17 MCP-exposed tools
  • JSON-RPC protocol
  • Pre-warmed tool cache
  • Render-ready (stateless)

30+ Tools, 7 Categories

Every tool is MCP-ready with full JSON Schema definitions for automatic discovery.

4
Data Fetch
3
Scraping
3
Transform
5
Notify
4
Action
4
Control
2
AI