Primary view showing the Simulated Alexa+ Assistant (MCP-powered) and 3 sample email buttons for testing.
Users can paste messy client emails or use sample buttons to quickly auto-fill the text area.
Google Gemini AI extracts 4 tasks, identifying the client name, converting deadlines, and assigning priorities.
Clicking "Remind" generates a context-aware, polite follow-up email draft using the Gemini AI.
Confirmation alert after clicking send. Email dispatch is simulated for this demo environment
Users paste a client reply, and the AI analyzes the text to detect intent and suggest an action.
AI correctly identifies the task as "done" with high confidence and updates the MongoDB database.
Official MCP Inspector confirming the backend server connects and exposes the parse_email tool via JSON-RPC.

Inspiration: Freelancers waste hours managing client emails, tasks, and reminders manually. Existing tools are too complex.

What it does: It uses AI to extract tasks from messy emails, drafts polite client reminders, and automatically updates task statuses when clients reply. It also features an MCP server for Alexa+ integration.

How we built it: React for frontend, Node.js/Express for backend, MongoDB Atlas for database, Google Gemini API for AI, and MCP SDK for the Alexa+ track. Challenges we ran into: AWS Bedrock registration was blocked due to ID issues, so we pivoted to Google Gemini. Setting up the MCP server was also a great learning curve.

Accomplishments that we're proud of: Building a fully working, live full-stack AI application from scratch and successfully integrating the MCP protocol.

What we learned: How to build AI agents, how the MCP protocol works, and full-stack deployment using Render and Vercel.

What's next for AI Ops Copilot: Adding real email sending (via Resend) and making the Alexa+ voice skill live on actual devices.