Project Overview

The Goal

This project deploys an open-access AI application utilizing a custom dataset of humanitarian documents. The user flow supports immediate guest usage, with an optional authentication path for session persistence. The model is optimized to assist remote field staff by appending educational context to generated responses.

The Solution

This project utilizes a modern, low-overhead stack designed for high impact. Built on a Nuxt UI Pro template, the frontend offers immediate public access with an optional login path bridged to baena.ai. The backend architecture avoids traditional server bloat by routing API requests directly to n8n automation workflows, which handle AI model interaction and data persistence via Supabase

Key Objectives

• Develop a public, session-based chat experience for anonymous users with a clear call-to-action to register.
• Implement a delegated authentication system for users to sign up/log in to save and retrieve their entire chat history.
• Build a robust backend integration connecting the Nuxt server to an n8n webhook and persisting all data in a Supabase database.

Audience & Stakeholders

• Primary Users: Humanitarians practicioners with the need of real time data analysis and decision-making support.
• Secondary Users: Humanitarian organizations looking to enhance their operational efficiency through AI-powered tools.
• Key Stakeholders: This is a solo project managed by Jesus Baena, who is responsible for all aspects of development, deployment, and maintenance.

The Plan & Key Features

Overall Approach

The project initially proved its concept via a no-code Bubble prototype. To ensure scalability and control, I am now migrating to a custom architecture using a Nuxt UI Pro template as the visual foundation. This process involves decoupling the template's default authentication and AI services to replace them with our specific requirements. Development follows a backend-first strategy: establishing the custom API and Supabase persistence layer before implementing the dynamic frontend logic

Core Components

• Authentication System: A delegated flow redirecting users to baena.ai for login/signup and returning them to the chat app with an active session.
• Core Chat Interface: A dynamic UI featuring a resizable message input, conversation history display, AI assistant selector, and a user menu that adapts based on authentication status.
• Backend Integration: A Nuxt server API that routes requests to a designated n8n webhook, passing the message, thread_id, and assistant_id, with all conversations saved to Supabase.

Timeline & Deliverables

Major Milestones

• [Date] : Project Kickoff & Requirements Finalized
• [Date] : Backend Integration (API, n8n, Supabase) Complete
• [Date] : MVP Deployed and Live at chat.baena.ai

Final Deliverables

• A live and fully functional web application accessible at chat.baena.ai.
• The final source code repository for the application.
• A presentation or document summarizing the features and architecture.