AI Day Planner

AI-powered day planning with semantic todo search. Built with FastAPI, LangChain, Ollama, and Next.js.

Features

• AI Chat Assistant: Conversational AI that helps you plan your day using Ollama's qwen3:latest model
• Semantic Todo Search: Find related todos using vector similarity search with PGVector
• Smart Suggestions: AI suggests new todos that you can accept or reject
• Day Math Tool: Calculate dates, time between events, and scheduling
• Real-time Streaming: Chat responses stream in real-time using SSE
• Full Todo Management: Create, edit, complete, and delete todos

Architecture

Frontend (Next.js + React + MUI)  <--->  Backend (FastAPI + LangChain)  <--->  Postgres + PGVector
     |                                              |
     +---------------- SSE Stream -------------------+

Prerequisites

• Docker and Docker Compose
• Ollama installed locally
• Node.js 20+ (for local frontend development)
• Python 3.11+ with uv (for local backend development)

Quick Start

1. Setup Ollama

Install and pull the required models:

# Install Ollama from https://ollama.com

# Pull required models
ollama pull qwen3:latest
ollama pull nomic-embed-text-v2-moe:latest

2. Start Services

# Start Postgres (port 5439), backend (port 8000), and frontend (port 3000)
docker-compose up -d

# Run database migrations
cd backend
uv sync
uv run alembic upgrade head

3. Access the Application

• Frontend: http://localhost:3000
• Backend API: http://localhost:8000
• API Docs: http://localhost:8000/docs

Development

Backend Development

cd backend

# Install dependencies
uv sync

# Run with hot reload
uv run python -m uvicorn app.main:app --reload

# Run migrations
uv run alembic upgrade head

# Create new migration
uv run alembic revision --autogenerate -m "description"

# Run tests
uv run pytest

Frontend Development

cd frontend

# Install dependencies
npm install

# Run with hot reload
npm run dev

# Generate OpenAPI client from backend
npm run generate-client

API Endpoints

Todos

• GET /todos - List all todos
• POST /todos - Create a new todo
• GET /todos/{id} - Get a specific todo
• PATCH /todos/{id} - Update a todo
• DELETE /todos/{id} - Delete a todo

Chat

• POST /chat/stream - Streaming chat endpoint (SSE)
• GET /chat/suggestions - Get pending suggestions
• POST /chat/suggestions/{id}/accept - Accept a suggestion
• POST /chat/suggestions/{id}/reject - Reject a suggestion

Environment Variables

Create .env files in both backend/ and frontend/ directories:

Backend (.env)

DATABASE_URL=postgresql+asyncpg://postgres:postgres@localhost:5439/planner
OLLAMA_BASE_URL=http://localhost:11434
CORS_ORIGINS=http://localhost:3000
DEBUG=false

Frontend (.env.local)

NEXT_PUBLIC_API_URL=http://localhost:8000

How It Works

Semantic Search

Todos are automatically embedded using nomic-embed-text-v2-moe:latest when created/updated. The agent can search semantically:

• "Find my gym-related tasks" → Matches "workout", "exercise", "fitness"
• "What about project X?" → Finds contextually related tasks
• "Any errands to run?" → Matches "grocery", "shopping", "dry cleaning"

Day Math

The agent uses the day_math tool for date calculations:

• "3 days from now" → Future date calculation
• "Days until Friday" → Countdown calculation
• "Last Monday's date" → Past date lookup

Suggestion Flow

1. Agent calls suggest_todo() with title, description, and reasoning
2. Suggestion appears as a card in the chat UI
3. User clicks Accept → Todo is created
4. User clicks Reject → Suggestion is removed

Project Structure

.
├── backend/
│   ├── app/
│   │   ├── main.py          # FastAPI app
│   │   ├── agent.py         # LangChain agent with tools
│   │   ├── models.py        # SQLAlchemy models
│   │   ├── schemas.py       # Pydantic schemas
│   │   ├── database.py      # Database connection
│   │   ├── config.py        # Settings
│   │   └── routers/
│   │       ├── todos.py     # Todo CRUD endpoints
│   │       └── chat.py      # Streaming chat endpoints
│   ├── alembic/             # Database migrations
│   ├── pyproject.toml       # uv dependencies
│   └── Dockerfile
├── frontend/
│   ├── src/
│   │   ├── app/             # Next.js App Router
│   │   │   ├── page.tsx     # Chat page
│   │   │   └── todos/
│   │   │       └── page.tsx
│   │   ├── components/
│   │   │   ├── Chat.tsx     # Main chat interface
│   │   │   ├── ChatMessage.tsx
│   │   │   ├── SuggestionCard.tsx
│   │   │   └── TodoList.tsx
│   │   ├── hooks/
│   │   │   └── useChat.ts   # SSE streaming hook
│   │   └── lib/
│   │       ├── api.ts       # API client
│   │       └── theme.ts     # MUI theme
│   ├── package.json
│   └── Dockerfile
└── docker-compose.yml

Troubleshooting

Ollama Connection Issues

If the backend can't connect to Ollama:

1. Ensure Ollama is running: ollama serve
2. Verify the OLLAMA_BASE_URL is set to http://localhost:11434

Database Issues

To reset the database:

docker-compose down -v
docker-compose up -d db
cd backend
uv run alembic upgrade head

Port Conflicts

• Postgres: 5439 (configured to avoid conflict with local Postgres)
• Backend: 8000
• Frontend: 3000

License

MIT