LlamaIndex

1
import os
2
import asyncio
3
from llama_index.tools.airweave import AirweaveToolSpec
4
from llama_index.core.agent.workflow import FunctionAgent
5
from llama_index.llms.openai import OpenAI
6
7
# Initialize the Airweave tool
8
airweave_tool = AirweaveToolSpec(
9
    api_key=os.environ["AIRWEAVE_API_KEY"],
10
)
11
12
# Create an agent with the Airweave tools
13
agent = FunctionAgent(
14
    tools=airweave_tool.to_tool_list(),
15
    llm=OpenAI(model="gpt-4o-mini"),
16
    system_prompt="""You are a helpful assistant that can search through
17
    Airweave collections to answer questions about your organization's data.""",
18
)
19
20
# Use the agent to search your data
21
async def main():
22
    response = await agent.run(
23
        "Search the finance-data collection for Q4 revenue reports"
24
    )
25
    print(response)
26
27
if __name__ == "__main__":
28
    asyncio.run(main())

1
from llama_index.tools.airweave import AirweaveToolSpec
2
3
airweave_tool = AirweaveToolSpec(api_key="your-key")
4
5
# List collections
6
collections = airweave_tool.list_collections()
7
print(f"Found {len(collections)} collections")
8
9
# Simple search
10
results = airweave_tool.search_collection(
11
    collection_id="finance-data",
12
    query="Q4 revenue reports",
13
    limit=5
14
)
15
16
for doc in results:
17
    print(f"Score: {doc.metadata.get('score', 'N/A')}")
18
    print(f"Text: {doc.text[:200]}...")

1
result = airweave_tool.advanced_search_collection(
2
    collection_id="finance-data",
3
    query="Q4 revenue reports",
4
    limit=20,
5
    retrieval_strategy="hybrid",
6
    temporal_relevance=0.3,
7
    expand_query=True,
8
    interpret_filters=True,
9
    rerank=True,
10
    generate_answer=True,
11
)
12
13
documents = result["documents"]
14
if "answer" in result:
15
    print(f"Generated Answer: {result['answer']}")

1
answer = airweave_tool.search_and_generate_answer(
2
    collection_id="finance-data",
3
    query="What was our Q4 revenue growth?",
4
    limit=10,
5
    use_reranking=True,
6
)
7
print(answer)  # "Q4 revenue grew by 23% to $45M compared to Q3..."

1
# Keyword search for exact term matching
2
results = airweave_tool.advanced_search_collection(
3
    collection_id="legal-docs",
4
    query="GDPR compliance",
5
    retrieval_strategy="keyword",
6
)
7
8
# Neural search for semantic understanding
9
results = airweave_tool.advanced_search_collection(
10
    collection_id="research-papers",
11
    query="papers about transformer architectures",
12
    retrieval_strategy="neural",
13
)
14
15
# Hybrid search (default) - best of both worlds
16
results = airweave_tool.advanced_search_collection(
17
    collection_id="all-docs",
18
    query="machine learning best practices",
19
    retrieval_strategy="hybrid",
20
)

1
results = airweave_tool.advanced_search_collection(
2
    collection_id="news-articles",
3
    query="AI breakthroughs",
4
    temporal_relevance=0.8,  # 0.0 = no recency bias, 1.0 = only recent matters
5
)

1
airweave_tool = AirweaveToolSpec(
2
    api_key="your-api-key",
3
    base_url="https://your-airweave-instance.com",
4
)

1
from llama_index.llms.ollama import Ollama
2
3
agent = FunctionAgent(
4
    tools=airweave_tool.to_tool_list(),
5
    llm=Ollama(model="llama3.1", request_timeout=360.0),
6
)