Max Peng

Max Peng

Menu


Projects I have worked on


Emolytics

Emolytics - Helps video creators know and understand how users are responding to their videos

Tools used to make project: React, Redux, React Router, Postgress, AWS, Express, Microsoft Emotions Api, Webgazer

Oversaw a team of 5 in executing the vision and maintaining an efficient workflow

Captured and reformatted images of the user through the webcam for eye tracking and emotional analysis data from the WebGazer module and Microsoft’s Emotions API to show to video creators.

Increased overall data management efficiency by over 70% by remapping the data calculations from the client to SQL’s highly optimized math functions.

Incorporated React Router and Redux to create a seamless and responsive user experience

Link
 

Learning With Lesssons

Learning with Lessons - Platform for teachers and students to create and access course material in an organized manner

Tools used to make project: React, React Router, Mongoose, Express, C3, Passport, Socket.io

Used Socket.io to create real time chat rooms associated with individual lesson pages for students to ask questions

Created C3 and D3 charts to help lesson creators visualize lesson popularity

Managed and troubleshooted git workflow for a 4 man team


 

Kiva Impact

Kiva Impact - Help loaners quantify the impact of their loans

KPCB Sponsor Prize at Cal Hacks 2.0
Tools used to make project: Bootstrap, Google Charts, Quandl

Service built at Cal Hacks 2.0 where we helped loaners quantify the impact of their loans in the community of which they are loaning to.

Created impact algorithm used to quantify the effect of loans in the community. Utilized data from the World Bank and the United Nations to determine the "power", importance and social implication of the loan.

Github Repo
 

Agrilerts

Agrilerts - Help farmers optimize crop production

Finalist at Big Red Hacks
Tools used to make project: Bootstrap, Forecast.io, Twilio, Python, Flask

Service for farmers in the developing world to receive updates on weather and planting tips to increase crop yields through text messages. Using a technology that the farmers do have to supplement their lack of internet

Coded the text message commands and responses to be intuitive and cost effective for farmers. Designed the introduction website and implemented Forecast.io into the backend.

Github Repo
 

Vibe

Vibe - Predict the emotion of a region from social media images

Tools used to make project: Microsoft Project Oxford Computer Vision API, OpenCV's ORB algorithm, Microsoft Azure Machine Learning, TwitterSearch, mapbox.js, Bootstrap, Adobe Illustrator

Web-app that analyes the emotions of a region in the world from images uploaded to social media. Mined images from Twitter and evaluated emotional responses to the image. After aggregated emotional response in a region, we produced a vizualization of the different moods.

Helped design the web-app as well as weight the input variables of the incoming picture to what emotion it generates.

Deployed at https://radiant-plateau-7115.herokuapp.com/