lilakelland Goto Github PK

Hi there and welcome to my github page!

I love working with data and playing with code to solve problems. I learned the fundamentals of software development through projects that smooth small anoyances in daily life, then jumped into data with the Data Science Bootcamp at Lighthouse Labs.

The 3 primary projects on the go are:

Ravelry Recommender Engine

(Repository: Ravelry-Recommender-Engine

• Designed a recommender system using data pulled from the knitting website Ravelry’s API in order to help reduce decision paralysis when choosing from their over 600k patterns.
• Used Sklearn and Implicit in Python to leverage item-item, content and collaborative filtering techniques.
• Deployed as a Flask app on AWS.

UnBurnt

(Repositories: UnBurnt-iOS-App, UnBurntArduino, UnBurnt-Sensor-Client and UnBurnt-REST-API)

• To prevent burnt BBQ food, sensors (thermocouple and flame) are attached to an arduino with wifi to send data via python API to mongoDB/ iOS app.
• The iOS app receives push notifications when it's time to check the food, if it's too hot, on fire and when it's too cold.
• Uses state machine and background modes, so you never have to turn on or off system, it'll automatically turn on when reached min temp and off when cooled back down.
• Currently setting up supervised machine learning to determine burning point (from temp slope and temp), but am going to need much more data (and resolder my initial sensors).
• This project is spread among 3 repositories:
• The UnBurnt-REST-API repository contains the Flask-RESTful python API code. Docker scripts (to run python/ SQL on raspberry pi server), SQL schema setup and project UML diagram are also found here.
• The UnBurnt-Sensor-Client repository contains the code that reads and processes the sensor information from the arduino and determines cooking state based on this information.
• The other 2 repositories (UnBurnt-Arduino and UnBurnt-iOS-App are more self explainitory containing the sensor schematic and Aruduino code and iOS code respectively)

Papaoutai

(Repositories: Papaoutai-REST-API, [Papaoutai-iOS-App]((https://github.com/LilaKelland/Papaoutai-iOS-App) and Papaoutai-Arduino

• An iOS app that tracks the time that adults in the house spend (aka hides) in bathroom.
• Uses BLE from an Arduino Nano to connect to iphone in background mode, and tracks time spent within a preset proximity.
• On BLE disconnection, the iOS app sends time data to Posgresql via python api.
• Day and Week usage is displayed along with week averages on a iPhone app (similar to the "screentime" app). Currently working on combining the proximity tracking and usage display apps into one.
• Next steps after this is to use alerts to update user with weekly usage stats and including time percentage increase or decrease.