Inspiration
I wanted to create the kind of study buddy we wish every student had. When students want to study, they often have to figure out what to study, break everything into tasks, create a schedule, and track their own progress before they can even get started. Many tools help with parts of this process, but the planning still falls on the student. I imagined Croc as a friendly companion who could help with both the planning and the studying, turning “I need to study” into a clear next step.
What it does
Crocademy is a web-based, Alexa+ inspired study companion built around Croc, a friendly AI assistant. Students can manage study tasks, organize subjects and topics, track their confidence, monitor progress, and schedule focus sessions in one workspace. Instead of only answering questions, Croc can understand study requests and use tools to help update the student's planner and related data. If the AI is unavailable, a keyword-based fallback keeps basic assistance working.
How we built it
I built Crocademy as a web application using React and TypeScript for the frontend and Express for the backend. I integrated Amazon Bedrock Converse with Nova 2 Lite to power Croc's conversational responses. I also built an MCP server that exposes study-related tools, allowing the assistant to interact with tasks, plans, and other study data. A shared backend state keeps updates consistent across the planner and other screens. The experience is designed to demonstrate an Alexa+ style assistant through a web interface rather than an actual Alexa device integration.
Challenges we ran into
One of my main challenges was connecting natural-language conversations to actions that actually change the application. I wanted Croc to do more than generate helpful-sounding responses, so we had to connect the AI, MCP tools, backend state, and frontend into one consistent experience. I also needed a fallback so students could still receive basic assistance when the AI service could not respond. Balancing the conversational experience with a functional, reliable planner was an important part of the process.
Accomplishments that we're proud of
I am proud that Crocademy brings study planning, task management, focus sessions, and progress tracking into one connected experience. Croc is more than a chatbot sitting beside a dashboard: the assistant can use tools to interact with the same study data students see throughout the application. I am also proud to have combined Amazon Bedrock and MCP in a project focused on a relatable, everyday problem for students.
What we learned
I learned that building a useful AI assistant involves much more than connecting a model to a chat interface. The assistant needs tools, shared state, and a reliable way to turn a student's request into an action. I also learned the importance of fallback behavior and designing the experience around what users actually need, rather than adding AI just for the sake of having it.
What's next for Crocademy
I want to make Croc an even more helpful study companion by improving personalized study recommendations, helping students break large goals into manageable daily tasks, and adapting plans based on progress and topic confidence. I'd also like to expand voice interactions, add more ways to celebrate milestones, and explore integrations that help students move from planning to focused learning with less friction. Our long-term goal is to make starting to study feel easier, less overwhelming, and a little less lonely.
Built With
- bedrock
- express.js
- mcp
- nova
- react
- tailwind
- vite
- webspeechapi
Log in or sign up for Devpost to join the conversation.