Hi, my name is Ken Wu
I'm a Student, Developer, and Problem Solver.

Get in touch

About Me

Profile Image

I'm Qianghao Ken Wu, a dedicated Computer Science and Business Administration student at the University of Waterloo and Wilfrid Laurier University.
My academic journey is marked by a passion for Data Science, Machine Learning, Software Development, and Supply Chain Management, with hands-on experience garnered through roles at Laurier Fintech, Keywords AI, and Delphai.
Skilled in Python, Tensorflow, Keras, spaCy, SQL, and C, and fluent in English and Chinese, I navigate seamlessly through technical and multicultural environments, bringing innovation and analytical skills to the forefront of technology and business.
Let’s connect if you're as enthusiastic as I am about leveraging data and technology to drive actionable insights and solutions!

View Resume

Projects

Sovestico

- Engineered "Sovestico," an ethical investment web app, integrating React, Vite, Python, and APIs (OpenAI, Cohere, Tavily, Yahoo Finance) for dynamic ESG data analysis and stock simulation
- Designed an intuitive, responsive UI with React and Vite, simplifying ESG insights for users and promoting informed, value-aligned investment decisions
- Implemented real-time financial data fetching and natural language stock transactions, enhancing user experience and engagement in sustainable investing

Source Code
Project Image

LeaseEase - Winner of 2024 McHack Telus Environment and Social Sustainable Future Prize

- Developed LeaseEase, a tool empowering vulnerable groups in Canada's housing crisis by simplifying legal processes using a blend of LLM, RAG, and user-friendly interfaces.
- Integrated technologies like Streamlit, OpenAI, Cohere, and ChromaDB, overcoming complex backend-frontend integration and data processing challenges.
- Focused on expanding LeaseEase's functionalities and collaborating with government bodies, aiming to enhance social justice and accessibility of legal resources.

Source Code
Project Image

MedChat - Winner of 2023 Cohere RAG Challenge

- Integrated Cohere's Classify API to determine user intent, triggering specific models like the brain tumor CNN for accurate diagnosis.
- Utilized Cohere Embedding and RAG, MedChat efficiently generates accurate citations from over 1000 WebMD pages.
- Streamlit is used in MedChat for instantaneous data streaming from Cohere API and the pre-trained CNN model, enhancing user interaction.

Source Code
Project Image

DirectU - Winner of Hack the North 2023 Best Use of Cohere Prize

-Developed the backend of DirectU W, a course recommendation system, utilizing the Cohere API for text embedding to calculate Similarity Score and Text Summarizer as well as a chat feature for analysis
- Developed backend using Flask and Cohere Embedding API, achieving a 99% reduction in loading time
- Collaborated in a team of 4 on a full-stack web app with React, MongoDB, and Selenium

Source Code
Project Image

Lichess and Chess.com Github Readme Badges

- Developed an automated attendance tracking system for classrooms
- Utilized a tech stack consisting of Python, Linux, OpenCV, and the face-recognition library
- Programmed a raspberry pi to scan students' faces automatically

Source Code
Project Image

Animal Crossing and Doom Classifier

- Created a deep-learning model to classify images as animal crossing or doom related
- Successfully deployed the model in a local docker container
- Demonstrated proficiency in deep learning, computer vision, and deployment

Source Code
Project Image

London Bike Usage Regressor

- Utilized Python programming language to build a regression model predicting the number of bikes rented in London
- Implemented different machine learning algorithms such as XGBoost, RandomForest, and Linear Regression for model development
- Incorporated various factors such as temperature, wind speed, time, humidity, holiday, weather code, and season for feature selection and model training
- Enabled bike providers to optimize the distribution of bikes for maximum profit and usage

Source Code
Project Image

Facial Recognition Attendance Taker

- Developed an automated attendance tracking system for classrooms
- Utilized a tech stack consisting of Python, Linux, OpenCV, and the face-recognition library
- Programmed a raspberry pi to scan students' faces automatically

Source Code
Project Image

Contact

Got Any Questions?

Shoot Me an Email