README.md

metadata

title: Filo - DRM focused Construction Planner
emoji: 🏗️
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 6.0.0
app_file: app.py
pinned: false
tags:
  - mcp-in-action-track-enterprise
  - mcp-in-action-track-customer
  - disaster-risk-management
  - philippines
  - geospatial
  - hazard-assessment

🏗️ Filo - DRM focused Construction Planner

An AI-powered multi-agent system that leverages Model Context Protocol (MCP) to plan disaster-resistant construction in the Philippines. This project demonstrates enterprise-grade MCP integration across specialized agents for risk assessment, research, cost analysis, and facility mapping.

📺 Watch Demo Video | 💼 LinkedIn Post | 𝕏 X Post

🎯 MCP Integration Highlights

This project showcases Model Context Protocol as the backbone for agent communication and data access:

MCP-Powered Agents

• 🌋 Risk Assessment Agent - Uses Philippines Disaster Risk MCP for real-time hazard data (seismic, volcanic, hydrometeorological)
• 🔍 Research Agent - Integrates DuckDuckGo MCP and Fetch MCP for construction standards and best practices
• 💰 Cost Analysis Agent - Leverages Perplexity MCP for up-to-date material pricing and market data
• 📍 Facility Locator Agent - Uses Google Maps API to identify critical infrastructure (hospitals, schools, emergency services)
• 🎨 Visualization Agent - Generates context-aware building visualizations using Gemini API with image generation

MCP Architecture Benefits

• Standardized Communication: All agents communicate via MCP protocol for consistent data exchange
• Modular Design: Each agent is independently deployable with its own MCP integrations
• Real-time Data Access: MCP servers provide live access to disaster risk databases, web search, and market data
• Scalable Orchestration: Blaxel platform manages agent lifecycle and MCP connections

✨ Features

• 🌊 Risk Assessment: Real-time analysis of seismic, volcanic, and hydrometeorological hazards via MCP
• 🏗️ Construction Recommendations: AI-generated building-specific guidance for disaster resistance
• 💵 Cost Analysis: Dynamic material cost estimation using MCP-powered market data
• 🗺️ Facility Mapping: Interactive maps showing nearby critical infrastructure
• 🖼️ Visual Generation: AI-generated building visualizations based on construction plans

How to Use

1. Select Building Type: Choose from residential, commercial, industrial, institutional, or mixed-use
2. Enter Coordinates: Provide latitude and longitude within Philippine bounds (4°N-21°N, 116°E-127°E)
3. Specify Building Area (optional): Enter estimated building area in square meters
4. Generate Plan: Click "Generate Construction Plan" button
5. Wait for Analysis: The system collects data from multiple MCP servers and generates comprehensive analysis (takes approximately 1-2 minutes)
6. Review Results: Navigate through tabs to view different aspects of the plan

Example Locations

• Manila: 14.5995°N, 120.9842°E
• Cebu: 10.3157°N, 123.8854°E
• Davao: 7.1907°N, 125.4553°E

🛠️ Technology Stack

Core Technologies

• Model Context Protocol (MCP) - Standardized agent communication and data access
• Blaxel Platform - Multi-agent orchestration and MCP server hosting
• Gradio 6 - Modern web interface framework
• LangChain - MCP adapters and LLM integration
• OpenAI API - GPT-4 for agent reasoning and synthesis
• Gemini API - Image generation for building visualizations

MCP Servers Used

1. Philippines Disaster Risk MCP - Custom MCP server for Philippine hazard data
2. DuckDuckGo MCP - Web search for construction research
3. Fetch MCP - Content retrieval for standards and regulations
4. Perplexity MCP - Real-time market data for cost analysis

Architecture

• 5 Specialized Agents: Risk Assessment, Research, Cost Analysis, Facility Locator, Visualization
• 1 Orchestrator Agent: Coordinates multi-agent workflow with LLM-powered synthesis
• HTTP-based Communication: RESTful APIs for agent-to-agent communication
• MCP Protocol: Standardized data access across all external services

🏆 Why This Project Stands Out

Enterprise-Grade MCP Implementation

• Custom MCP Server: Built a specialized Philippines Disaster Risk MCP server for geospatial hazard data
• Multi-Agent Orchestration: Demonstrates MCP at scale with 5+ agents working in concert
• Real-World Impact: Addresses critical disaster risk management needs in the Philippines
• Production Ready: Deployed on Hugging Face Spaces with full CI/CD pipeline

Customer Value

• Life-Saving Insights: Helps builders create safer structures in disaster-prone areas
• Cost Optimization: Balances safety requirements with budget constraints
• Comprehensive Planning: Integrates risk, research, costs, and logistics in one platform
• Accessible Interface: User-friendly Gradio UI makes complex analysis accessible to non-technical users

⚠️ Disclaimer

This tool provides guidance only. Always consult with licensed engineers, architects, and local authorities for actual construction projects.