Skip to main content

Building a Virtual Workplace Assistant

This tutorial series demonstrates how to create an Virtual Workplace Assistant using CAP and LangGraph, with website crawling, data retrieval, and deployment capabilities.

Tutorial Series​

Part 1: Website Crawling & Data Ingestion​

Learn how to set up website crawling using Apify to gather content for your Virtual Workplace Assistant.

Key Steps:​

  • Setting up Apify crawler
  • Configuring crawl patterns
  • Data preprocessing
  • Storage preparation

Part 2: Data Retrieval Setup​

Implement efficient data retrieval using Milvus DB for the Virtual Workplace Assistant.

Implementation:​

from cap.agents import VirtualWorkplaceAgent
from cap.storage import MilvusStorage

class WorkplaceAssistant(VirtualWorkplaceAgent):
    def __init__(self):
        super().__init__()
        self.vector_store = MilvusStorage()
        
    async def retrieve_context(self, query: str) -> list:
        return await self.vector_store.similarity_search(query)

Part 3: AWS Deployment​

Deploy your Virtual Workplace Assistant as an embeddable chatbot on AWS.

Deployment Configuration:​

module "workplace_assistant" {
  source = "./modules/lambda"
  
  name = "virtual-workplace-assistant"
  runtime = "python3.9"
  handler = "handler.lambda_handler"
  
  environment = {
    MILVUS_HOST = var.milvus_host
    APIFY_TOKEN = var.apify_token
  }
}

Features​

  • Website content crawling
  • Vector-based retrieval
  • Conversation memory
  • AWS Lambda deployment
  • Embeddable widget

Try It Out​

  1. Clone the repository
  2. Follow setup in Getting Started
  3. Configure Apify and Milvus DB
  4. Deploy using provided Terraform configuration
  5. Embed the widget in your website

Learn More​