LockIn

LockIn WatchOS app
LockIn WebApp for Mac, Windows Notifications

We were inspired to build Lock In when we realized that traditional tools like timers and to‑do lists only remind you after you’ve already drifted.

Our LockIn System:

Apple Watch > Firestore > Cloud Functions + ML > Desktop Notifications (both Mac & Windows)

We wanted something that senses distraction in the moment and brings you back before you lose momentum.

Throughout this project, we learned how to process physiological signals extracting time‑domain HRV features like mean NN, SDNN, and RMSSD from raw ECG and BVP data. We also gained experience merging two very different datasets (the WESAD baseline recordings for “focused” states and a Kaggle cognitive load dataset for “distracted” states), cleaning and labeling them, and building an end‑to‑end machine learning pipeline in VS Code.

We go to learn the hard way about the pain points that Apple HealthKit and had to improvise and look at creating supplementary health fields based on data we were able to get from it.

Our goal was ambitious and this hackathon helped us realize a glimpse of what a non privacy invasive productivity tracker + motivator could be. Thanks to Google's cloud offerings - building a server less architecture for our LockIn system helped various of our team members work on different stacks and bring it to life.

We hope that we can make LockIn much more robust, so that productivity enthusiasts like ourselves can build cross platform use cases to boost productivity during study & work.