DealFlow is an advanced sales agent framework built on CAMEL (Communicative Agents for "Mind" Exploration of Large Language Model Society). It provides a robust system for creating autonomous, intelligent sales agents capable of conducting natural, contextually-aware sales conversations with potential customers.
DealFlow aims to replicate and enhance the capabilities of SalesGPT while leveraging the powerful agent communication framework provided by CAMEL. The system enables sales agents to:
- Conduct natural, contextually aware sales conversations
- Analyze and adapt to different stages of the sales process
- Utilize tools for product searches, payment processing, scheduling, and more
- Maintain conversation history and learn from previous interactions
- Generate effective responses tailored to customer needs
DealFlow is built using the following components from the CAMEL framework:
-
Core Agent System
- Uses CAMEL's ChatAgent for the primary sales agent
- Leverages RolePlaying for more complex multi-agent scenarios
-
Conversation Management
- Implements LongtermAgentMemory for maintaining conversation context
- Stores and retrieves relevant conversation history
-
Sales Stage Management
- Custom stage analyzer for determining the current sales conversation stage
- Stage-appropriate response generation
-
Tools Integration
- Product knowledge base search
- Payment processing
- Calendar/meeting scheduling
- Email communication
-
Configuration System
- Customizable company information
- Adaptable salesperson profile and style
- Adjustable conversation goals and strategies
The main controller class that manages the sales agent's behavior, conversation flow, and tool usage.
A specialized component for determining the current stage of the sales conversation and suggesting appropriate strategies.
Manages the integration and execution of various tools that enhance the agent's capabilities.
External API for integrating DealFlow into applications, websites, or other systems.
from dealflow import DealFlow
from camel.models import ModelFactory
from camel.types import ModelPlatformType, ModelType
# Initialize a language model
model = ModelFactory.create(
model_platform=ModelPlatformType.OPENAI,
model_type=ModelType.GPT_4O_MINI,
)
# Configure the sales agent
sales_agent = DealFlow(
model=model,
salesperson_name="Alex Johnson",
salesperson_role="Solutions Consultant",
company_name="TechSolutions Inc.",
company_business="We provide enterprise-grade cloud computing solutions for businesses of all sizes.",
company_values="We believe in empowering businesses through technology that's accessible, reliable, and forward-thinking.",
conversation_purpose="understand the client's cloud infrastructure needs and present our scalable solutions",
use_tools=True
)
# Start the conversation
sales_agent.seed_agent()
# Human input handling
human_input = "Hi, I got your email about cloud solutions. Can you tell me more?"
response = sales_agent.step(human_input)
print(response)While inspired by SalesGPT, DealFlow offers several advantages through the CAMEL framework:
-
Advanced Agent Communication: CAMEL's specialized communication protocols enable more natural and effective interactions.
-
Enhanced Memory Systems: More sophisticated memory management for improved context retention.
-
Flexible Tool Integration: Streamlined approach to integrating external tools and APIs.
-
Scalable Multi-Agent Support: Better support for multi-agent scenarios where multiple sales representatives or specialists might be involved in the sales process.
-
Improved Reasoning Capabilities: CAMEL's focus on agent reasoning leads to more strategic sales conversations.
pip install dealflow- Python 3.9+
- camel-ai
- Required API keys (depending on the model and tools used)