From install to a grounded answer in about ten lines. You need an API key first; if you do not have one yet, see Your API key.
pip install searchcandy
export SEARCHCANDY_API_KEY=sc_live_...
from searchcandy import SearchCandy
sc = SearchCandy()
# point it at a folder of .txt or .md files
sc.ingest_folder("./contracts")
r = sc.query("What is the termination notice period?")
print(r.quotes[0].text) # the exact passage
print(r.quotes[0].source_id) # the file it came from
t = sc.query("What does clause 99.9 say?") # not in the docs
print(t.not_covered) # True. No guess, no filler.The first ingest builds your graph, which takes a moment. A small folder is ready in seconds; queries after that return in well under a second.
Your documents became a knowledge graph with a saved version 1. Answers come with sources you can check, and a question the documents cannot answer was turned away honestly. That is the whole product in four calls.
Next: Adding documents →