Hey there! This is the code for a company I was building for the last six months :D
🎯The aim of Crysta was to help decrease student procrastination by tracking their energy levels. Basically, Crysta would track various different metrics throughout the day – things like the number of distracting sites visited, the number of task switches, and the amount of time spent per task – and then, using an algorithm (the data_handler.py file under the flask branch), compile all these metrics into a single number known as an energy score.
🔍Crysta would then look at how these scores changed throughout the day to help determine when you were most likely to be productive, and when you were most likely to be creative. Later on, I even integrated some features that would make it easier to work – such as a tasks list and a pomodoro timer.
💡The original idea for the company came after reading Chris Bailey’s Hyperfocus, a book I’ve talked about in a couple other of my videos. As the author puts it, we each have a unique Biological Prime time – where we’re most productive and energetic. Similarly, we also have a time where we’re least energetic, but more creative. Crysta aimed to be a solution that would measure your energy levels – your biological prime time - and then personalize your daily schedule around these findings Theoretically, this would mean significantly higher productivity for the end user.
You can find out more (and use Crysta yourself) at our website: https://tinyurl.com/Crysta-MVP
While the company itself didn't take off, I learned a LOT through the experience. Here's a quick article with all the insights, lessons, and frameworks I've aquired through this venture - and, specific action items to avoid future mistakes. https://aditya-dewan124.medium.com/my-first-company-failed-heres-what-i-learned-and-what-you-should-avoid-5174ee78481a
Here's a quick, 3min video talking about this! https://www.loom.com/share/8003c48133cb47aa9fc4b43e359eb036
There are 2 parts to Crysta:
- 🌐A webapp - made with the MERN stack (MongoDB, Express, React, and Node.js) and deployed on Heroku (link on the website)
- ⚡A desktop energy tracker - an API made with Python and Flask that starts a local server and feeds the energy levels into the webapp.
Few people have tried tracking someone's energy levels directly from their desktop activity. But, after reaching out to some professors - namely, Piers Steel Peter Gröpel, the writers of a mega-trial responsible for finding procrastination factors - I decided to make this happen. Once we had enough data, the idea was to apply Artifical Intelligence (MLPs) to more accurately determine this. (here's the paper the two professors wrote: https://www.researchgate.net/publication/222569814_A_mega-trial_investigation_of_goal_setting_interest_enhancement_and_energy_on_procrastination)
What made this app particularly challenging to build, especially for a first-time web-developer like me, was that the webapp had to communicate with desktop app. Crysta’s desktop app is what collected the necessary data, like task-switches and distracting sites, to generate an energy score – which then was posted to the database. The idea was that, the desktop app (which was made with Python and Flask) would start tracking your energy levels the second you logged into the webapp – meaning that there had to be repeated communication going on between the two apps. **So, to accomplish this, I learned the ins-and-outs of APIs, requests, servers, and how all of these different pieces work and come together to create a good user experience. **
On the other hand, working on features like the pomodoro timer and the task manager also taught me the basics of database management and storage using the CRUD (create, read, update, and delete) method, as well as MongoDB queries and best practices.
All in all, this repo is almost 100 thousand lines of total code (and 1.7k+ lines of written code)!
There are three branches:
- staging (this is the main branch -> where all new features were pushed) - this version can be found at https://crysta-staging.herokuapp.com/
- flask_energy_tracker (the code for the Python/Flask API + the energy algorithm) - this can be found at https://drive.google.com/file/d/1d1Fw2UBuvYmu9M9CJrHf_iRkNXMUVO6p/view?usp=drivesdk
- the development branch (for when features have been tested and deemed good for production https://crysta-app.herokuapp.com/
This project was AWESOME - from designing the figma mockups to coding for days on end, I wouldn't have traded this for anything else.
Thanks for reading! I hope this will be useful for a LOT of people (and advancing research).
