LazyFlow AI: Turning Messy Messages into Actionable Tasks with Gemma 3

Description and use

LazyFlow AI automatically extracts tasks, deadlines, and priorities from long messages using a local Gemma 3 4B model via Ollama, enabling users to edit, save, and track task completion without external AI services.

Problem

Parents and others often receive long, unstructured messages (e.g., school announcements, workplace emails) that contain multiple reminders, deadlines, and activities, making it hard to identify and organize actionable tasks.

Outcome

LazyFlow AI reduces manual effort in extracting tasks from lengthy messages by automating task identification, deadline and priority extraction, and task grouping, while enabling editing, completion tracking, and collection saving—all locally without external AI services.

Workflow

• Paste message into the web app
• Frontend sends text to backend via HTTP
• Backend forwards prompt + message to Ollama API
• Ollama runs Gemma 3 4B locally and returns structured JSON
• Backend validates JSON and returns tasks to frontend
• Frontend displays tasks in dashboard
• User edits, deletes, or marks tasks as completed
• User can group tasks, select subsets, and save into named collections
• Collections and tasks are persisted in browser localStorage
• User tracks completion progress per collection

Evidence

LazyFlow AI is a simple, full‑stack AI‑powered task management application that transforms long, messy, or unstructured messages into clear, actionable tasks using Google's open‑weight Gemma 3 4B model running locally through Ollama.

Source

When a user pastes a message and clicks Generate Tasks, the React frontend sends the text to the Express backend.

Source

The backend validates the response before returning the extracted tasks to the frontend.

Source

For this MVP, I used localStorage to persist tasks and collections across browser refreshes without introducing a database or authentication system.

Source