GPT4All - Extract Information from PDF Bank Statements into JSON

Extract Data from Bank Statements (PDF) into JSON files with the help of GPT4All / Llama3 LLM

• List PDFs from your drive that roughly have a similar layout and you expect an LLM to be able to extract data in a systematic way

• Formulate a concise prompt (and instruction) and try to force the LLM to give back a JSON file with always the same structure (Mistral seems to be very good at that)

• Use GPT4All wrapper to put document and query before the LLM

• Collect the responses

• Extract the data from JSON files, either with the help of Regex or just convert the JSON with KKNIME nodes

• Make sure they have the same structure
  
  => you need to have GPT4All installed and a suitable model downloaded to your "gpt4all_models" folder. You can choose the model then in the component

External resources

• Medium: Llama3 and KNIME — Build your local Vector Store from PDFs and other Documents
• GPT4All - Run Large Language Models Locally
• Download the whole LLM workflow group in order to get the folder
• KNIME Spaces: KNIME for Generative AI
• GitHub - A Streamlit App to convert your files into Vector Stores and chat with them with LLMs
• Medium - Chat with local Llama3 Model via Ollama in KNIME Analytics Platform - Also extract Logs into structured JSON Files
• Medium - Ollama and Llama3 - A Streamlit App to convert your files into local Vector Stores and chat with them using the latest LLMs