Jose Castillo

Private map channel system where users can tag points of interest and invite others to their channels. Google Maps, AngularJS, Android, Firebase

- Mobile app that suggested Spotify playlists by user’s personality and emotion with Watson and Naïve Bayes classifiers.
- Led programming team of 4 by overseeing the design and implementation of the app.
- Built and trained Bayes classifier to quickly categorize songs from their content and provide meaningful matching.
- Mathematically matched personality from analyzing user’s Twitter account against playlists ranked by customized sentiment analysis over playlist’s… Show more

- Mobile app that suggested Spotify playlists by user’s personality and emotion with Watson and Naïve Bayes classifiers.
- Led programming team of 4 by overseeing the design and implementation of the app.
- Built and trained Bayes classifier to quickly categorize songs from their content and provide meaningful matching.
- Mathematically matched personality from analyzing user’s Twitter account against playlists ranked by customized sentiment analysis over playlist’s songs.

Technologies: Machine Learning, NodeJS, Python, MongoDB, Ionic. Show less

- Built hand gesture tool to send SMS with Twilio API by typing Morse code and selecting suggested words with hands movements through a NodeJS server connected with Web Sockets.
- Created custom Myo gesture to select words by mathematically handling the gyroscope and the accelerometer vectors.

Technologies: NodeJS, Socket.io, JavaScript, Myo Armband.

- Mobile app to connect boaters to one another and capture reported hazards from boating crowd on the water.
- Awarded Trophy from FIU Business School Crowdsourcing competition for best crowd sourcing project design.

- Collaborated with team of 4 to create a mobile bidding app that allowed users to search local products by auto-generated tags using visual recognition.
- Built RESTful API on PHP (Slim Framework), loaded data into Parse through Python scripts and created a library to interact with the Clarifai API.

Technologies: PHP, Python, Machine Learning, Parse.

- Project lead on a team of 4 that maximized profit estimation of a mailing campaign of 13 million donors by 650% using consolidation trees and neural networks.
- Awarded Silver Medal from FIU Business School Data Mining competition for 2nd best predictive model.

Technologies: SAS, Data Mining.

Redesigned core business process (operations control). Performed To-Be simulations to verify improvement. Achieved enhancement of performance by 77% through mapping, and alignment of processes & technology. IBM Holosofx, BPM.

Created ERD model and OLAP star schema. Enabled insights for regional managers by devising visualizations and report dashboards. Microsoft SQL Server, Tableau, Excel, Visual Studio, ETL.

- Created desktop tool to measure news polarization in social media using a multithreaded K-Means clustering algorithm.
- Built a multithreaded RSS reader, which scraped and web crawled more than 3000 daily feeds from the world’s top news networks over 7 months.