I am currently exploring LangChain and find it really useful to contextualize LLMs for an enterprise and use it to build AI powered bots or solve common problems related to support or any process for a given enterprise.
However, I am not finding any good documentation on data privacy. Does anyone know if LangChain would upload all data to LLM vendor server if I have to use it's capabilities to build question and answer bot for my company's processes and confidential documents?
Here is code snippet on how I am using it -
import os
import sys
from langchain import PromptTemplate
from langchain.chains import RetrievalQA
from langchain.llms import OpenAI
os.environ["OPENAI_API_KEY"] = ""
llm = OpenAI(temperature=0.9) # model_name="text-davinci-003"
from langchain.document_loaders import TextLoader
# loader = TextLoader('state_of_the_union.txt', encoding='utf8')
loader = TextLoader('The_Chronology_of_Puranic_Kings_and_Rigv.txt', encoding='utf8')
from langchain.indexes import VectorstoreIndexCreator
index = VectorstoreIndexCreator().from_loaders([loader])
# query = "What did the president say about Ketanji Brown Jackson"
# query = "How much time elapsed from the beginning to the end of the Nanda dynasty"
# query = "When did MahÀpadma Nanda came to power"
for line in sys.stdin:
if 'q' == line.rstrip():
break
print(f'Input : {line}')
print(index.query(line))
print("Exit")
Will LangChain upload whole data ["The_Chronology_of_Puranic_Kings_and_Rigv.txt"] to openai in this case?