DealSense — Plataforma Multi-Agent para Análise de Imóveis Comerciais
selvaonline/realestate-ai-agent
Análise automática de deals imobiliários comerciais com raciocínio transparente e citações
Ver no GitHubMCP server que implementa uma plataforma de IA multi-agente especializada em análise de oportunidades de investimento em imóveis comerciais (CRE). Usa LangGraph.js com supervisor e 6 agentes especializados, integrando 20+ ferramentas de busca e análise, arquitetura modular de domínio, avaliações de nível forense e interface tipo Perplexity com raciocínio transparente.
Instalação
npx convez-ai selvaonline/realestate-ai-agent1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Triagem automática de oportunidades de investimento em CRE com critérios financeiros
- Análise financeira underwriting (cap rate, DSCR, ROI) de portfólios imobiliários
- Agregação de dados de múltiplas fontes (LoopNet, Crexi, brokers) com citações verificáveis
- Geração de relatórios e investment memos estruturados para análise por private equity
Veredito

Solução sofisticada e domain-specific para o nicho de análise de imóveis comerciais, com arquitetura de multi-agentes bem estruturada e ênfase em rastreabilidade de fontes. Ideal para equipes de PE/RE que precisam automatizar due diligence, mas requer investimento em setup e manutenção de integrações com portais de listagem.
README
A tradução para o português deste README ainda está na fila.
🏢 RealEstate AI Deal Agent
An intelligent agent that finds and analyzes commercial real estate investment opportunities using a Perplexity-style interface with transparent reasoning, source citations, and streaming answers.
##Demo Question Find retail centers under $20M, 8%+ cap, strong demographics, repositioning opportunity
Find retail centers under $20M, 8%+ cap, strong demographics, repositioning opportunity
Find medical office buildings or urgent care facilities, cap rate 7%+
✨ Features
🤖 Intelligent Agent Flow
- Transparent Thinking - See the agent's reasoning process in real-time
- Source Citations - Numbered references [1], [2], [3] for all information
- Streaming Answers - Progressive text generation with inline citations
- Multi-Source Search - Aggregates from LoopNet, Crexi, and broker sites
💬 AI Chatbot (NEW!)
- Conversational Interface - Ask questions about deals, scores, and analytics
- Explainability - "Why is the Risk Score 56?" - Get detailed breakdowns
- Artifact Generation - "Create an IC memo for deal #1" - Instant reports
- Portfolio Q&A - "How many Tier A deals?" - Query your results
- Quick Actions - Pre-configured prompts for common tasks
- Context-Aware - Knows about your current search results
👉 See CHATBOT_QUICK_START.md to get started!
📊 Property Analysis
- Automated Extraction - Property details, pricing, NOI, cap rate
- Financial Underwriting - DSCR, cash flow, ROI calculations
- Visual Confirmation - Screenshots of source listings
- Deal Cards - Clean presentation of investment opportunities
🎨 Modern UI
- Perplexity-Inspired Design - Clean, professional, engaging
- Progressive Disclosure - Thinking → Sources → Answer → Deals
- Responsive Layout - Works on desktop and mobile
- Dark Theme - Easy on the eyes with blue accents
🚀 Quick Start
Prerequisites
- Node.js 18+
- OpenAI API key
- Serper API key (for web search)
Installation
# Clone the repository
git clone https://github.com/selvaonline/realestate-ai-agent.git
cd realestate-ai-agent
# Setup backend
cd orchestrator
npm install
cp .env.example .env
# Edit .env with your API keys
# Setup frontend
cd ../deal-agent-ui
npm install
Running
# Terminal 1: Start backend (port 3001)
cd orchestrator
npm run dev
# Terminal 2: Start frontend (port 4200)
cd deal-agent-ui
ng serve
# Open browser
open http://localhost:4200
📖 Documentation
- Perplexity Flow Overview - Detailed explanation of the flow
- Flow Diagrams - Visual sequence and state diagrams
- Developer Guide - How to extend and customize
- Before/After Comparison - See the improvements
🎯 How It Works
1. Query Understanding
User: "Find multifamily deals in Dallas with cap rate > 6%"
Agent: 🔍 Understanding your query...
2. Source Discovery
Agent: 🔍 Searching commercial real estate listings...
Sources found:
[1] Crexi - Multifamily Property in Dallas
[2] LoopNet - 123 Main Street Investment
[3] Broker Site - Dallas Multifamily Portfolio
3. Data Extraction
Agent: 🔍 Analyzing property listings...
🌐 Navigating to crexi.com...
📸 [Screenshot captured]
4. Answer Synthesis
Found a promising listing [1]: **Crexi Multifamily Property**
located at 123 Main St, Dallas, TX. The asking price is
$2,500,000. Net Operating Income (NOI) is $200,000.
The cap rate is 8.00%. DSCR is 1.45. • • •
5. Source Attribution
Sources:
[1] Crexi - Multifamily Property Dallas
123 Main Street, Dallas, TX investment opportunity
https://www.crexi.com/property/...
🏗️ Architecture
┌─────────────┐
│ Angular │ ← UI Layer (Perplexity-style interface)
│ UI │
└──────┬──────┘
│ SSE Stream
▼
┌─────────────┐
│ Express │ ← API Layer (Event orchestration)
│ Server │
└──────┬──────┘
│ Function Calls
▼
┌─────────────┐
│ LangChain │ ← Agent Layer (Decision making)
│ Agent │
└──────┬──────┘
│ Tool Invocations
▼
┌─────────────┐
│ Tools │ ← Tool Layer (Search, Browse, Finance)
│ Layer │
└─────────────┘
🛠️ Tech Stack
Backend
- Node.js + TypeScript - Runtime and type safety
- Express - HTTP server
- LangChain - Agent orchestration
- Playwright - Browser automation
- OpenAI GPT-4 - Language model
- Serper - Web search API
Frontend
- Angular 17 - Framework with signals
- TypeScript - Type safety
- Server-Sent Events - Real-time streaming
- Responsive CSS - Modern design
📊 Example Output
Thinking Steps
🔍 Understanding your query...
🔍 Searching commercial real estate listings...
🔍 Analyzing property listings...
Answer with Citations
Found a promising listing [1]: **Vista Ridge Apartments** located
at 5847 Forest Ln, Dallas, TX 75230. The asking price is $16,500,000.
Net Operating Income (NOI) is $1,251,642. The cap rate is 7.58%.
DSCR is 1.52.
Sources
[1] Crexi - Vista Ridge Apartments
Dallas multifamily investment opportunity...
https://www.crexi.com/property/...
🔧 Configuration
Environment Variables
# orchestrator/.env
OPENAI_API_KEY=sk-... # OpenAI API key
SERPER_API_KEY=... # Serper search API key
OPENAI_MODEL=gpt-4o-mini # Model to use
PORT=3001 # Server port
# Optional: Browser debugging
BROWSER_HEADED=false # Show browser window
BROWSER_ENGINE=chromium # chromium | webkit
BROWSER_DEVTOOLS=false # Open DevTools
Customization
See Developer Guide for:
- Adding new event types
- Customizing thinking messages
- Styling the UI
- Adding new tools
🐛 Troubleshooting
Events not showing in UI
- Check browser console for errors
- Verify SSE connection in Network tab
- Ensure backend is running on port 3001
No search results
- Verify
SERPER_API_KEYis set - Check query format (be specific)
- Try broader search terms
Browser automation failing
- Install Playwright browsers:
npx playwright install - Try different engine:
BROWSER_ENGINE=webkit - Enable headed mode for debugging:
BROWSER_HEADED=true
🤝 Contributing
Contributions welcome! Please read our Developer Guide first.
📄 License
MIT License - see LICENSE file for details
🙏 Acknowledgments
- Inspired by Perplexity AI
- Built with LangChain
- Powered by OpenAI GPT-4
Made with ❤️ for commercial real estate investors