I'm Kaijuan Xing(KX). I'm a master student at UT Austin iSchool, with an interest in Data Science and Machine Learning. If you found yourself talking a lot about "Virtual Assistant", "AI", "Chatbot", "NLP", "Cognitive Science", I'm with you! If you like "Metallica", "Iron Madien", then I like you! I am a punky girl and have some fun facts: 1. I use a different name for each Starbucks store 2. I love the smell of gas and paint.
I want to be a Data Scientist with particular interest in NLP to power human-computer
conversation or interaction. I have been involved in some chatbot development projects during my internship at AT&T, where I participated several different chatbots with different function like "schedule vacation" and "technician workflow optimization" and etc. I utilized the skills I learnt from iShcool like "Data Wrangling", "Database Management", "Machine Learning" and "Metadata generation". These skills all give me better understanding of the problem arised from big data and how to solve it by process.
Buid and Design chatbot evaluation metrics to monitor chatbot performance(intents understanding, conversation completion), chatbot development project management(time and cost saving), chatbots functioning status('up' bot, number of user interacted with) Developed chatbot dashboard for internal usage analysis, built unsupervised machine learning model to group bot functions.
Developed chatbots in Python using SDKs from Amazon Lex, IBM Watson and Common Machine Learning Platform. Developed Django program to train Natural Language Processing (NLP) models applying CMLP microServices and deployed production pipeline hosted on Kubernetes.
Build a model that’s capable of detecting different types of of toxicity like threats, obscenity, insults, and identity-based hate for online social media's comments. The model is based on RNN and char-ngram.
View ProjectAn anoymous secrets sharing app targeted at UT students. Users are free to explore posts around them and comment on any post as being anoymous in the 'Incognito' way.
View ProjectDeveloped a book/author/publisher information searching web application hosted on Google Cloud Platform using Flask. Processed raw JSON data to Postgres database using Flask Sqlalchemy and programmed unit tests to test database models and deployed on Travis CI. View Project
Implemented WebCrawler to extract metadata information about question/questioner, answer/answerer. Pre-processed data for generating features and trained model using Logistic Regression to test answer rating at open question/answer community
View Project