kumardatascience/Multi-Source-RAG-AI-System-with-Query-Routing

1"""Tiny workflow demo to understand the pattern. Not used by the chatbot."""2 3import asyncio4from llama_index.core.workflow import (5    Workflow,6    step,7    Event,8    StartEvent,9    StopEvent,10)11 12# 1. Define our custom event types (just data containers)13class GreetingEvent(Event):14    message: str15 16 17class ShoutingEvent(Event):18    message: str19 20 21# 2. Define the workflow with its steps22class DemoWorkflow(Workflow):23 24    @step25    async def greet(self, ev: StartEvent) -> GreetingEvent:26        """Step 1: receives the start, emits a greeting."""27        print("👋 Step 1: greeting...")28        return GreetingEvent(message=f"Hello, {ev.name}!")29 30    @step31    async def shout(self, ev: GreetingEvent) -> ShoutingEvent:32        """Step 2: takes the greeting, shouts it."""33        print("📢 Step 2: shouting...")34        return ShoutingEvent(message=ev.message.upper())35 36    @step37    async def finish(self, ev: ShoutingEvent) -> StopEvent:38        """Step 3: ends the workflow with the final result."""39        print("🏁 Step 3: done.")40        return StopEvent(result=ev.message)41 42 43# 3. Run it44async def main():45    wf = DemoWorkflow()46    result = await wf.run(name="Mihir")47    print(f"\nFinal result: {result}")48 49 50if __name__ == "__main__":51    asyncio.run(main())