DayBuddy AI

Inspiration

Staying organized throughout the day can be harder than it sounds. Tasks pile up, priorities get confusing, and it becomes difficult to balance productivity with breaks and focused work. I wanted to build something that could make daily planning simpler and help people turn a long to-do list into an actionable schedule.

That idea led to DayBuddy AI — an AI-powered daily planning assistant designed to help users organize tasks, prioritize important work, plan their day, and stay focused.

What It Does

DayBuddy AI combines task management, AI-powered planning, and focus tools in one simple web application.

• Task Management: Add, complete, search, filter, and sort tasks, with data saved locally in the browser.
• AI Daily Planner: Uses Gemini to generate a structured daily plan based on tasks and available working hours.
• Visual Schedule: Displays planned activities in an easy-to-understand timeline, including priorities and breaks when included in the generated plan.
• Focus Mode: Provides a countdown-based focus session to help users concentrate on a selected task.
• Progress Dashboard: Shows task statistics and progress throughout the day.
• MCP Tools: Includes a Model Context Protocol (MCP) server with tools for prioritizing tasks and generating a time-based daily schedule.

How I Built It

I built DayBuddy AI using HTML, CSS, and JavaScript for the frontend, with Node.js and Express powering the backend. I integrated Google's Gemini API to generate daily plans and kept the API key in an environment variable instead of exposing it in frontend code.

For the Alexa+ hackathon track, I added a separate MCP server using the official Model Context Protocol TypeScript ecosystem's JavaScript SDK packages. It exposes two tools: prioritize_tasks and create_daily_plan. I then created an MCP client test script to connect to the server, discover its tools, call them, and verify their responses.

I also used Git and GitHub to version-control the project and publish its source code with setup instructions.

Challenges I Faced

One challenge was integrating AI-generated planning into an existing task management application while keeping the schedule structured and easy to display. I needed to handle API requests, parse structured responses, and account for tasks that could not fit into the available time.

Another challenge was learning how MCP works in practice. I created a separate MCP server so the existing website could continue running independently, then tested actual client-to-server tool calls over Streamable HTTP. Resolving Git configuration and repository merge issues was another part of getting the project ready to share.

What I Learned

This project helped me understand how to connect a frontend, backend, and AI API into a working application. I learned more about environment variables, API integration, structured data, task scheduling, Git workflows, and testing an MCP server with a real client.

Most importantly, I learned that building an AI application is not just about generating intelligent responses. It is also about making those responses useful, predictable, and accessible through a clear user experience.

What's Next

I would like to connect DayBuddy's task data more directly to its MCP tools, improve schedule generation with smarter break and time-conflict handling, and explore how the experience could work with Alexa+ through an appropriate integration or a clearly demonstrated web-based simulation.

My goal is to make DayBuddy AI a practical daily companion that helps people spend less time organizing their work and more time getting it done.

Built With

• ai
• css
• express.js
• generative-ai
• github
• google-gemini-api
• html
• javascript
• model-context-protocol-(mcp)
• node.js