CrewAI

pip install crewai langchain-tensoras

export TENSORAS_API_KEY="tns_your_key_here"

from langchain_tensoras import ChatTensoras
 
llm = ChatTensoras(model="llama-3.3-70b")

from crewai import Agent, Task, Crew
from langchain_tensoras import ChatTensoras
 
llm = ChatTensoras(model="llama-3.3-70b")
 
# Define agents
researcher = Agent(
    role="Research Analyst",
    goal="Find accurate, up-to-date information on the given topic.",
    backstory="You are an experienced research analyst who excels at finding and synthesizing information.",
    llm=llm,
    verbose=True,
)
 
writer = Agent(
    role="Technical Writer",
    goal="Write clear, concise technical content based on research findings.",
    backstory="You are a skilled technical writer who turns complex topics into accessible content.",
    llm=llm,
    verbose=True,
)
 
# Define tasks
research_task = Task(
    description="Research the current state of retrieval-augmented generation (RAG) technology. Focus on recent advances, common architectures, and best practices.",
    expected_output="A structured research summary with key findings.",
    agent=researcher,
)
 
writing_task = Task(
    description="Write a blog post about RAG technology based on the research findings. The post should be 500 words, technically accurate, and accessible to developers.",
    expected_output="A polished blog post in markdown format.",
    agent=writer,
)
 
# Create and run the crew
crew = Crew(
    agents=[researcher, writer],
    tasks=[research_task, writing_task],
    verbose=True,
)
 
result = crew.kickoff()
print(result)

from crewai import Agent
from langchain_tensoras import ChatTensoras
 
# Use a larger model for complex reasoning
senior_analyst = Agent(
    role="Senior Analyst",
    goal="Perform deep analysis of complex data.",
    backstory="You are a senior data analyst with 15 years of experience.",
    llm=ChatTensoras(model="deepseek-r1"),
)
 
# Use a faster model for simpler tasks
formatter = Agent(
    role="Report Formatter",
    goal="Format analysis results into clean reports.",
    backstory="You are a report formatting specialist.",
    llm=ChatTensoras(model="llama-3.3-70b"),
)

from crewai import Agent, Task, Crew
from crewai.tools import tool
from langchain_tensoras import ChatTensoras
 
llm = ChatTensoras(model="llama-3.3-70b")
 
@tool("Search Knowledge Base")
def search_kb(query: str) -> str:
    """Search the company knowledge base for information."""
    from tensoras import Tensoras
 
    client = Tensoras()
    response = client.chat.completions.create(
        model="llama-3.3-70b",
        messages=[{"role": "user", "content": query}],
        knowledge_bases=["kb_a1b2c3d4"],
    )
    return response.choices[0].message.content
 
support_agent = Agent(
    role="Customer Support Agent",
    goal="Answer customer questions accurately using the knowledge base.",
    backstory="You are a helpful support agent with access to company documentation.",
    llm=llm,
    tools=[search_kb],
)
 
support_task = Task(
    description="Answer the following customer question: 'How do I configure SSO for my organization?'",
    expected_output="A clear, step-by-step answer based on the knowledge base.",
    agent=support_agent,
)
 
crew = Crew(agents=[support_agent], tasks=[support_task])
result = crew.kickoff()
print(result)

crew = Crew(
    agents=[researcher, writer],
    tasks=[research_task, writing_task],
    process="sequential",  # default
)

crew = Crew(
    agents=[researcher, writer],
    tasks=[research_task, writing_task],
    process="hierarchical",
    manager_llm=ChatTensoras(model="llama-3.3-70b"),
)