Technology Stack:

• Frontend: Built with React to provide a responsive, browser-based web application.

• Backend: Powered by a Python API to handle logic and system communication.

• Core Intelligence: Utilizes the OpenAI API (Flash 4) for processing and generation.

Database Management:

• PostgreSQL: Serves as the primary database for structured data, including user accounts, client details, and onboarding checklists.

• VectorDB (Chroma): Acts as the "AI memory," storing vector embeddings of uploaded company documents.

AI Logic & Security:

• RAG Model: Implements Retrieval-Augmented Generation to ensure the AI only draws information from specific, admin-provided documents.

• Accuracy & Control: By using RAG, the system prevents hallucinations and ensures responses are strictly grounded in the company's verified data.

The three main features/functions of this system are the following: AI Chatbot, Adaptive Checklist, and Document Upload.

• The AI Chatbot assists new clients of a company complete onboarding. It can answer general questions and help clients fill out paperwork accurately.

• To ensure the client completes all documentation required, the Adaptive Checklists keeps the client on task, and gives a graphical view of their progress.

• Document Upload is where all documents completed by the client are submitted and uploaded to the company database.

By using a RAG model, this ensures that only accurate information is displayed in AI generated responses. By using end-to-end encryption, the system ensures that all data is securely transferred to and from the client and the company database.