Inspiration
My primary inspiration comes from Habitica, a gamified task list that I used to use over 6 years ago. My project is a spinoff of that idea but adding numerous additional integrations to improve user experience.
What it does
TaskPets is a gamified task/goals list that has built-in AI features powered by Gemini that makes your task list easier to manage. There are numerous features that it contains when creating a task or a goal.
- Subtask Generation: The AI will split up your big task into numerous subtasks that you can complete one by one to ensure progress towards the task
- Built-in Timer: You can see how long you spent on each task and feel productive after!
- Voice to Task Generation: You can use voice input powered by Deepgram to create tasks automatically. This feature works on both web and the mobile app.
For goals specifically:
- Completion Validation: Part of working towards a continuing goal is to be held accountable, the AI will ask for proof and evaluate if it meets the requirements
- History: Every completion is logged and could be seen at any time
A mobile (andriod) app was also created for easy task making on the go!
All information is stored in a MongoDB database and the finished product is deployed on Vercel.
How we built it
The frontend is primarily made html+css and the backend is express both of which is hosted on Vercel. The database that is used for this project is hosted on MongoDB. A good portion of the code was generated through prompting using Cursor and Github Copilot.
Challenges we ran into
The primary challenge was debugging when implementing the MongoDB database and the Vercel deployment. My method of editing cookies only worked locally and did not work once deployed on the web.
Accomplishments that we're proud of
This is the first successful AI integration, voice integration, and MongoDB backend that I have done before. The AI integration and voice integration were a lot easier than I thought but MongoDB did take me a good amount of time to figure out and debug all the API calls.
What we learned
Throughout the 24h I have learned a lot more about how to prompt and work with an AI Agent. I also learned how to structure my project a lot better to minimize token usage and be more readable for Agents
What's next for TaskPets
Improve one of the core features, the pets! Right now the sprite looks pretty bad and that would need to be overhauled to make it a more appealing process. There are also some visual bugs that have not been patched.
Log in or sign up for Devpost to join the conversation.