I have export postman’s API into JSON, and load into pinecone (refer attached image). Also confirmed the upladed type is “JSON” and content been uploaded to pinecone
i used 4 different AI modes which are smart enough to handle most cases
but the AI’s response is never accurate after many attempts.
Hi @weitat.bipo, welcome back to the community!
According to your case, because the json data is in Binary file, not from json data, you need to use Binary as the Type of Data in your Default Data Loader node.
To reproduce this, simply change the Type of Data from JSON to Binary. Thanks!
Thanks for the promptly response cute kitty 
assuming my JSON (binary) is 1.3MB (can’t reduce further), how could data loader to consume the file still? thanks
As what it said, you need to do some “advanced” solving, you need to split the data from your json and use loop node and add the loader there.
any template for me to references? 
thanks in million~
update: i might have found the split example… still figuring out. there ain’t forloop too
I don’t know this may help you or not, but you can check some of ready-to-use workflows from official n8n community workflows site
You can check some of the workflows here:
[1]: Document-Based Chatbot with Memory using OpenAI, Pinecone and Google Drive | n8n workflow template
[2]: AI-Powered Accounting Reports from Sabre EDI with GPT-4 and Pinecone RAG | n8n workflow template
May it helps!
things are quite fishy… i managed to chop my array and looped it (10 items in that array)
but when i loop, it run into infinite loop and keep index the json (as below), it’s keep consume the write… forcing me to delete my pinecone index. Otherwise i gotto sell my kidney… 
still not sure the root cause… but scary
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
