My Regulatory AI Assistant
Video demonstrating the results of creating a PDF-querying Q&A Chatbot for aerospace regulation using open-source, no-code tools. Utilized Flowise platform with interacting Nodes.
Today I’m super excited to show you in a short video it’s possible to build a Q and A Chatbot that queries PDFs using open-source with no-code tools.
In this instance I have used aerospace Regulatory Articles to make queries and summarise information as it was my assistant.
For this project, I have employed Flowise, a platform that utilizes interacting Nodes.
The specific Nodes I have used are:
OpenAI - wrapper around OpenAI LLM
Conversional Retrieval QA Chain - chain that integrates with a retrieval component and allows to configure input parameters and perform Q and A tasks.
Pinecone Upsert Document - the upsert operation writes vectors into a namespace. It’s a way to manage information by adding new pieces of information or updating old ones.
Open AI Embeddings - it measures the relatedness of text strings.
Pdf File - it decodes the data from the PDF document and then loads the PDF content.
Recursive Character Text Splitter - it splits document data recursively to keep all paragraphs, sentences then words together as long as possible.
Ready to see AI in action? 🤖 Check out the below video:
About Flowise:
Flowise is an open-source tool that allows developers to create LLM applications without coding. It's a drag-and-drop UI that's built on top of LangChainJS. Flowise is designed to make it easy to visualize and build LLM apps.
The user guide is available through FlowiseAI.
Data used in the trial:
For this trial I have used the UK MAA Continuing Airworthiness Engineering (CAE) 4000 Series Regulatory Articles in PDF but you can use any PDF using the same foundations.
Disclaimer: The information provided in the newsletter and related resources is intended for informational and educational purposes only. It does not constitute professional advice, and any actions taken based on the content are at the reader's discretion.