Building Conversational Agents

Creating agents that maintain context in conversations

Conversational agents remember what you’ve said and maintain context throughout the chat.

---

Quick Start

from praisonaiagents import Agent

# Create a conversational agent with memory
agent = Agent(
    instructions="You are a friendly assistant. Remember what users tell you.",
    memory=True  # Enable conversation memory
)

# Have a conversation
agent.start("Hi! My name is Alex.")
agent.start("What's my name?")  # Agent remembers: "Alex"

Enable memory=True to let agents remember conversations.

---

How Memory Works

Feature	Without Memory	With Memory
Context	Each message is new	Remembers past messages
Personalization	None	Remembers preferences
Multi-turn	Limited	Full conversation flow

---

Use Cases

Customer Support

Remember customer details and issue history

Tutoring

Track learning progress and adapt

Personal Assistant

Remember preferences and tasks

Companion

Build relationship over time

---

Example: Customer Support Bot

from praisonaiagents import Agent

support_bot = Agent(
    name="SupportBot",
    instructions="""You are a customer support agent.
    
    - Greet customers warmly
    - Ask for details about their issue
    - Provide step-by-step solutions
    - Remember their name and issue details""",
    memory=True
)

# Multi-turn conversation
support_bot.start("Hi, I can't log into my account")
support_bot.start("My email is [e-mail]")
support_bot.start("I tried resetting but no email came")
# Agent remembers all context and provides relevant help

---

Example: Learning Tutor

from praisonaiagents import Agent

tutor = Agent(
    name="MathTutor",
    instructions="""You are a patient math tutor.
    
    - Explain concepts step by step
    - Use simple language
    - Give practice problems
    - Remember what the student has learned""",
    memory=True
)

# Teaching session
tutor.start("Can you help me with algebra?")
tutor.start("What does 2x + 5 = 13 mean?")
tutor.start("So I subtract 5 first?")
tutor.start("Then divide by 2 to get x = 4?")
# Tutor tracks progress and builds on previous explanations

---

Complete Example

from praisonaiagents import Agent

# Personal assistant with memory
assistant = Agent(
    name="PersonalAssistant",
    instructions="""You are a personal assistant.
    
    Remember:
    - User's name and preferences
    - Tasks they mention
    - Topics they're interested in
    
    Be helpful, friendly, and personalized.""",
    memory=True
)

# Build a relationship over multiple messages
assistant.start("Hi! I'm Sarah and I love hiking.")
assistant.start("What outdoor activities would you recommend?")
assistant.start("I prefer mountains over beaches.")
assistant.start("What should I do this weekend?")
# Assistant remembers Sarah likes hiking and mountains

---

Best Practices

Enable Memory

Use memory=True for any multi-turn conversation

Clear Instructions

Tell the agent what to remember in instructions

Consistent Personality

Define a clear personality in instructions

Handle Context

Agent automatically manages conversation history

---

Next: Building Research Agents

Learn how to create agents that search and gather information.