Add the Docs MCP to your AI coding tool to get help building with Respan. No API key needed.{
"mcpServers": {
"respan-docs": {
"url": "https://respan.ai/docs/mcp"
}
}
}
What is LanceDB?
LanceDB is a serverless vector database built on the Lance columnar format. It runs embedded in your application with no server to manage.
Setup
Install packages
pip install respan-tracing lancedb openai
Set environment variables
export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"
export OPENAI_API_KEY="YOUR_OPENAI_API_KEY"
Initialize Respan and query LanceDB
Respan auto-instruments LanceDB — search and table operations are captured as spans.from respan_tracing import RespanTelemetry
from respan_tracing.decorators import workflow, task
import lancedb
from openai import OpenAI
# Initialize — auto-instruments LanceDB
telemetry = RespanTelemetry()
client = OpenAI()
db = lancedb.connect("~/.lancedb")
docs = [
"Respan provides observability for LLM applications.",
"Traces capture the full lifecycle of an LLM request.",
]
def get_embeddings(texts: list[str]) -> list[list[float]]:
response = client.embeddings.create(model="text-embedding-3-small", input=texts)
return [item.embedding for item in response.data]
# Create table with embeddings
vectors = get_embeddings(docs)
table = db.create_table(
"docs",
data=[
{"text": docs[i], "vector": vectors[i]}
for i in range(len(docs))
],
mode="overwrite",
)
@task(name="search_docs")
def search_docs(query: str):
query_vector = get_embeddings([query])[0]
results = table.search(query_vector).limit(3).to_list()
return [row["text"] for row in results]
@workflow(name="rag_pipeline")
def rag_pipeline(query: str):
context = search_docs(query)
response = client.chat.completions.create(
model="gpt-4o-mini",
messages=[
{"role": "system", "content": f"Context: {chr(10).join(context)}"},
{"role": "user", "content": query},
],
)
return response.choices[0].message.content
result = rag_pipeline("How does tracing work?")
print(result)
View your trace
Open the Traces page to see LanceDB operations as spans in your trace tree. Configuration
LanceDB is auto-instrumented via Instruments.LANCEDB. No additional configuration is needed.
LanceDB tracing is currently available in the Python SDK only.
