AI Todo App

I wanted to build more than just another todo app—I wanted to explore the intersection of productivity tools and AI assistance. The goal was to create a task management application that could not only track todos but also provide intelligent assistance through an integrated AI chatbot.

Built the foundational task management features using TypeScript for type safety and maintainability. Users can create new todos, mark them as complete, and delete tasks they no longer need. The interface is clean and intuitive, focusing on getting things done without unnecessary complexity.

The standout feature is the AI chatbot integration. Instead of relying on cloud-based AI services, I set up Ollama to host a large language model locally on my machine. This approach offers privacy benefits and eliminates API costs while still providing powerful AI capabilities.

Chose Google's Gemma 3 as the AI model for its balance of performance and efficiency. Gemma 3 runs smoothly on local hardware while providing helpful, contextual responses. The model can assist users with task planning, breaking down complex projects, and general productivity advice.

Developed REST API endpoints that bridge the frontend application with the locally-hosted Ollama instance. This involved handling request/response cycles, managing conversation context, and ensuring smooth communication between the user interface and the AI backend.

This project deepened my understanding of local AI deployment, API design, and full-stack TypeScript development. Running AI models locally opened my eyes to the possibilities of privacy-focused AI applications—no data leaves the user's machine, yet they still get intelligent assistance.