WhatACharger ⚡
Community-Powered EV Charging Discovery with Conversational AI
💡 Inspiration
Electric vehicle drivers can experience range anxiety when they struggle to find a suitable charging point, especially in unfamiliar locations. We wanted to explore how community-powered charging discovery and conversational AI could make this experience simpler.
What if drivers could discover charging options offered by local hosts and ask an AI assistant for help instead of navigating multiple screens?
That idea inspired WhatACharger.
🚀 What It Does
WhatACharger is an EV charging discovery platform designed to connect drivers with charging options listed by local hosts.
Our platform brings together charger discovery, pricing and compatibility information, host connectivity, booking workflows, and an Alexa+-style conversational assistant experience.
✨ Key Features
- Charger Discovery Dashboard: Browse charger listings and explore available location, pricing, and vehicle compatibility details.
- Community-Powered Charging: Support discovery of charging opportunities offered by local hosts, businesses, and other charger owners.
- Host Dashboard: Provide workflows for hosts to add and manage charger listings.
- Booking Workflows: Support booking requests and related management functionality.
- Conversational AI Experience: Allow users to express charging needs in natural language through an assistant interface.
- MCP-Powered Tools: Expose structured charging operations through a Model Context Protocol server.
🧠 How We Use MCP
A key technical component of WhatACharger is our Node.js and Express MCP server, built using the Model Context Protocol SDK.
The server exposes four tools:
findChargers— Search charger listings.checkAvailability— Check the availability information supported by the server.getHostContact— Retrieve available host contact information.createBooking— Create a booking through the implemented workflow.
These tools provide a structured interface for compatible AI applications to interact with our charging functionality. We tested the tools using MCP Inspector.
Our goal is to demonstrate how an MCP-based architecture can connect conversational experiences with practical, domain-specific operations.
🛠️ How We Built It
We developed the WhatACharger web application using React and Vite for the frontend. We built the MCP server with Node.js, Express, and the Model Context Protocol SDK.
The frontend provides the user-facing charger discovery and host workflows, while the MCP server exposes tools for charger searches, availability checks, host contact retrieval, and booking creation.
We used MCP Inspector to connect to the server and test the implemented tools.
🧩 Challenges We Faced
One of our main challenges was connecting a user-facing EV charging experience with a structured tool-based architecture.
We worked on configuring the MCP server, establishing Streamable HTTP connectivity, exposing the tools, and testing their behavior through MCP Inspector. We also worked to bring the charging discovery workflow and conversational assistant concept into a cohesive experience.
This process helped us better understand how MCP servers expose capabilities to AI applications and how individual tools can support a larger application workflow.
📚 What We Learned
Building WhatACharger helped us explore full-stack development, API design, MCP tool integration, and the relationship between conversational AI and real application functionality.
We learned that an effective AI experience requires more than a chat interface: it needs clearly defined tools, structured inputs and outputs, and reliable underlying application logic.
🌍 What's Next
We envision expanding WhatACharger with more comprehensive charger coverage, improved location-aware discovery, stronger verification of charger availability, and deeper conversational assistant integration.
Our long-term goal is to make charging options easier to discover and help reduce range anxiety through a more connected EV charging ecosystem.
WhatACharger — Find a charger. Connect with confidence. Keep moving.
Built With
- artificial-intelligence
- css
- express.js
- html
- javascript
- mcp
- model-context-protocol
- node.js
- react
- rest
- vite
- web-development

Log in or sign up for Devpost to join the conversation.