Smart Agenda is a cross-platform task management application built with React Native (Expo) on the frontend and Node.js (Express) on the backend, using TypeScript across both. It is a personal project developed to practice and improve programming skill, using a modern and well-structured full-stack architecture, while integrating to AI providers.
The app is designed to support natural language input, transforming user messages into structured tasks using AI.
- ✅ Create, update, delete, and retrieve tasks
- 💬 Chat-style UI for intuitive task interaction
- 🧠 Natural language understanding powered by OpenAI (other sources planned)
- 📱 Web and Mobile experience (Android, iOS via Expo)
- 🌐 Backend API with structured request/response handling
- 🔐 Token-based authentication (under development)
- 🗂️ Organized architecture with clear separation of concerns
smart-agenda/
├── backend/
├── frontend/
└── README.md
- Node.js ≥ 18
- Expo CLI
- npm or yarn
- AndroidStudio (optional)
git clone https://github.com/FernandaCastro/smart-agenda.git
cd smart-agendacd backend
cp .env.example .env # Add your OpenAI key and other config
npm install
npm run devcd frontend
cp .env.example .env # Add your your IP to it
npm install
npx expo start
//or npx expo run:androidCreate a .env file based on .env.example in both frontend/ and backend/ folders as needed.
📈 Roadmap
-
Store tasks in a database
-
Implement user authentication
-
Add notification/reminder support
| AI Integration Level | Description | Status |
|---|---|---|
| 0 – Wrapper | Single call to LLM (prompt ➜ response) | ✔ consolidated |
| 1 – Tool-calling | LLM chooses and fills functions | ⚙️ in deployment |
| 2 – Orchestration | Think ➜ act ➜ observe cycle with 1–2 tools and small RAM memory | 🔜 next step |
| 3 – Full Agent | Planning, multiple tools, long memory, self-reflection, cost/error control | ⏳ 2–3 sprints away |
| 4 – Multi-Agents | Delegation between specialized agents, coordination, high-level goals | optional future |
This is a personal and non-commercial project created with the goal of practicing modern programming techniques and improving full-stack development skills. It also explores the integration of AI for natural language processing and interaction.
MIT License.