This is a RAG based genAI application that can retrieve relevant products from the databse that meet user"s requirements. This repository contains a GenAI-powered application built using LangChain as the main framework and Streamlit for GUI development. The LLM model used is AZURE openAI model, but user can freely use their preferred model. Vector database used int his project is PineCone.
- Product Search Tool: Enter the necessary information, the tool recommends suitable ADI products that meet the customer's needs.
- Cross Reference Search: Upload a competitor's datasheet, the tool suggests ADI products with the same or similar specifications.
To run the app locally, follow these steps:
-
Clone this repository:
git clone https://github.com/Anh-Do_adi/ChocoMango cd ChocoMango
-
Install dependencies:
pip install -r requirements.txt
-
Input your AZURE openAI key, AZURE embedding model API key, pinecone API key, pinecone index to .env file
-
Run the app:
streamlit run .\homepage.py
- Enter your query in the form. Only the Application/Purpose field is required, other fields are optional. But the LLM's accuracy can be imrpoved if these fields' information is provided.
- Analog Assist processes your input and generates a PST.
Field | Description |
---|---|
Application/Purpose | Describe your intended use case |
Category | Select a product category from the dropdown list |
Min Voltage | Specify the minimum operating voltage |
Max Voltage | Specify the maximum operating voltage |
Supply Current Max Limit | Define the maximum allowable supply current |
Min Temperature Range | Set the minimum operating temperature |
Max Temperature Range | Set the maximum operating temperature |
Package | Define the preferred package type |
Other Requirements | Additional constraints or preferences |
- THe user uploads a competitor's datasheet.
- Analog Assist analyzes the datasheet and provides top 2 most comparables products.
- You can also download the corresponding ADI datasheet simultaneously.
The following diagram illustrates the workflow of the application:
- Python 3.11.x
- OpenAI API Key (LLM: gpt-4o-mini, Embedding: text-embedding-3-large)
- Pinecone Vector DB