A sophisticated multi-agent CRM system integrated with Elasticsearch and Elastic Agent Builder. Manages customers, sales, support, and marketing through specialized agents leveraging semantic search and RAG capabilities on CRM data stored in Elasticsearch.

This platform use OpenAI SDK for routing between agents, verifying the user's intent and delegating the request to the appropriate agent.

┌──────────────────────────────────────────────────────────────────────┐
│                    CRM Orchestrator                                  │
│  ┌─────────────┬─────────────┬─────────────┬───────────────────────┐ │
│  │  Agent      │  Task       │  Customer   │  Global               │ │
│  │  Registry   │  Queue      │  (Elastic)  │  State (to implement) │ │
│  └─────────────┴─────────────┴─────────────┴───────────────────────┘ │
└──────────────────────────────────────────────────────────────────────┘
         │              │              │
         ▼              ▼              ▼
┌────────────────────────────────────────────────────────────────┐
│              Elastic Agent Builder                             │
│  ┌─────────────┬─────────────┬─────────────┬─────────────────┐ │
│  │  Semantic   │  RAG        │  Context    │  Tools          │ │
│  │  Search     │  Retrieval  │  Building   │  Integration    │ │
│  └─────────────┴─────────────┴─────────────┴─────────────────┘ │
└────────────────────────────────────────────────────────────────┘
         │
         ▼
┌────────────────────────────────────────────────────────────────┐
│                    Elasticsearch Cluster                       │
│  ┌──────────────┐ ┌──────────────┐ ┌──────────────┐            │
│  │ crm_customers│ │ crm_tickets  │ │ crm_         │            │
│  │              │ │              │ │ interactions │            │
│  └──────────────┘ └──────────────┘ └──────────────┘            │
│  ┌──────────────┐                                              │
│  │ crm_campaigns│                                              │
│  └──────────────┘                                              │
└────────────────────────────────────────────────────────────────┘

• Agent Registry: Manages registration and discovery of all agents
• Task Queue: Routes requests to appropriate agents based on OpenAI routing
• Customer Management: Centralized customer data access via Elasticsearch
• Global State (to implement): Maintains conversation context and agent states

• Semantic Search: Full-text search with fuzzy matching across all CRM indices
• RAG (Retrieval Augmented Generation): Context-aware responses using Elastic data
• Context Building: Automatic context assembly for LLM queries
• Tools Integration: Native integration with Elasticsearch tools

• Sales Agent: Lead qualification, opportunity analysis, customer scoring
• Support Agent: Ticket management, knowledge base search, issue resolution
• Marketing Agent: Campaign analysis, customer segmentation, performance metrics

• Python 3.10+
• Elastic Cloud (serverless project)
• OpenAI API key (for routing requests to agents)

# Create virtual environment
python -m venv .venv

# Activate (Windows)
.venv\Scripts\activate

# Activate (Linux/Mac)
source .venv/bin/activate

pip install -r requirements.txt

# Copy environment template
copy .env.example .env  # Windows
cp .env.example .env    # Linux/Mac

Edit .env with your configuration:

OPENAI_API_KEY=your_openai_api_key_here

# Model configuration
LLM_MODEL=gpt-4o
TEMPERATURE=0.7


# Elastic Cloud Configuration
ELASTIC_HOST=
KIBANA_HOST=
ELASTIC_CLOUD_API_KEY=

Runs a comprehensive demonstration with sample data:

python main.py --demo

Demo Features:

• Creates 8 sample customers with diverse profiles
• Generates 6 support tickets with various priorities
• Adds 8 customer interactions (email, calls, meetings)
• Creates 5 marketing campaigns with performance metrics
• Demonstrates all agent capabilities
• Shows advanced Elasticsearch aggregations
• Semantic search examples

Command-line interface with intelligent agent routing:

python main.py --chat

Chat Features:

• OpenAI-powered agent routing
• Conversation persistence per agent
• Real-time agent responses
• Commands:
  • /agents - List available agents
  • /exit - Exit chat

Example Chat Session:

crm> Show me enterprise customers with high engagement
[crm_sales_agent] Found 3 enterprise customers with engagement scores above 80:
1. Carlos López (CloudInc) - Score: 85, Status: Customer
2. Sofia Castro (HealthTech Solutions) - Score: 95, Status: Champion
3. Roberto Silva (Enterprise Solutions) - Score: 70, Status: Prospect

crm> What marketing campaigns are currently active?
[crm_marketing_agent] Currently active campaigns:
1. "Q1 Email Campaign - Enterprise Leads" - 38.8% open rate
2. "Webinar Series - Cloud Solutions" - 39.3% open rate

Modern web UI with markdown rendering:

uvicorn webapp.server:app --reload --port 8000

Web Features:

• Real-time chat interface
• Markdown-formatted responses
• Agent routing visualization
• Session management
• Responsive dark theme
• HTML rendering of tables, lists, and formatted text

Access: http://localhost:8000

from elastic import ElasticAgentBuilder, ElasticAgentConfig

config = ElasticAgentConfig(
    enable_semantic_search=True,
    enable_rag=True,
    top_k_results=10,
)

agent = ElasticAgentBuilder("custom_agent") \
    .with_config(config) \
    .with_repository("customers", CustomerRepository()) \
    .with_repository("tickets", TicketRepository()) \
    .with_description("Custom agent for specific tasks") \
    .build()

Customer(
    customer_id="uuid",
    name="John Doe",
    email="[email protected]",
    company="TechCorp",
    status="lead",  # lead, prospect, customer, inactive, champion
    lead_score=75,
    engagement_score=60,
    lifetime_value=15000.0,
    tags=["enterprise", "hot_lead"],
    custom_fields={"industry": "Software", "employees": 500},
    created_at=datetime.now(),
    last_contacted_at=datetime.now(),
)

Ticket(
    ticket_id="uuid",
    customer_id="customer-uuid",
    subject="Login issue",
    description="Cannot access account",
    status="open",  # open, in_progress, pending, resolved, closed
    priority="high",  # low, medium, high, critical
    category="access_management",
    channel="email",
)

Interaction(
    interaction_id="uuid",
    customer_id="customer-uuid",
    interaction_type="email",  # email, call, meeting, chat, social
    direction="inbound",  # inbound, outbound
    subject="Product inquiry",
    content="Customer interested in enterprise features",
    outcome="interested",
    next_action="schedule_demo",
)

Campaign(
    campaign_id="uuid",
    name="Q1 Email Campaign",
    campaign_type="email",
    status="active",  # draft, active, paused, completed
    target_segment="enterprise",
    sent_count=1000,
    opened_count=388,
    clicked_count=97,
    converted_count=15,
)

ELASTIC_CLOUD_ID=your_cloud_id:region
ELASTIC_CLOUD_API_KEY=your_api_key

1. Sales - Lead Qualification
2. Support - Ticket Resolution
3. Marketing - Campaign Analysis
4. Customer 360 View

POST /api/chat
Content-Type: application/json

{
  "message": "Show me enterprise customers with high engagement"
}

Response:
{
  "ok": true,
  "agent": "crm_sales_agent",
  "text": "Found 3 enterprise customers...",
  "html": "<h3>Enterprise Customers</h3><ul>...",
  "reasoning": "User is asking about enterprise customers..."
}

This project is licensed under the MIT License - see the LICENSE file for details.