Haaris Waleh Developer Page

Introduction

Hi, my name is Haaris Waleh. I am a Computer Scientist and recent graduate from UCSD. I specialize in Machine Learning and Artificial Intelligence, but I am also fascinated by Data Science, Physics, and Mathematics. I hope to use my skills and knowledge to make technology and innovations that can benefit others.



Experience

Here is a quick overview of some of my work experience. Refer to my resume or LinkedIn for additional details.

Data Scientist, Business Analytics- Freelance

  • Collected data from various web sources and processed for use.
  • Performed demand forecasting using a variety of time series analysis methods including ARIMA and lstm models.
  • Performed ensemble learning for prediction using different data sets.
  • Created reports and met with executives to discuss results and plans.



    • Machine Learning Intern- Bourbaki Group

  • Collaborated with a team of interns to develop a machine learning tool that predicts the output of organic chemical equations based on line diagram inputs.
  • Researched state of the art models and performed transfer learning with these models to allow the tool to work with hand drawn and computer generated inputs.
  • This tool was able to accurately recognize organic chemical structures whether they were hand drawn or computer generated.



    • Software Develoment Intern- Creatorsoft

  • Built an enterprise scale application for automating accounting.
  • Wrote middle tier code and performed testing of various components.



    • Projects

    Here are some of my recent projects. Please visit my GitHub for more.

    Real Time Text to Speech Translation App:
    Skills: Python, React Native, transformers, LLM, deep learning, transfer learning, fine tuning

    About: Simultaneously developing an Android and iOS app in React for real time text to speech translation. Translation will support multiple languages. App will be deployed, maintained, and updated as needed.


    Article Classifier:
    Skills: Python, sklearn, nltk, unsupervised learning, topic modeling, NLP, Pandas

    About: Python script that uses non-negative matrix factorization (NMF) to generate a topic modeler on a dataset of BBC articles to determine the main topic of each article in the dataset. This model can make predictions on user input files as well.


    Binary Image Segmenter:
    Skills: Python, PyTorch, U-Net, torchvision, Git

    About: U-Net pipeline for model that performs binary image segmentation on portraits of people. As team lead, I also created guides for my peers and set project goals. End result was presented to peers and faculty.

    Note: This project was conducted as part of a group project for the Association for Computing Machinery (ACM) club at UCSD.


    Digit Classification Optimization:
    Skills: Python, Pytorch, TensorFlow, CNN, MLP, logistic regression

    About: Performed optimization of digit classification models on the MNIST dataset. This included writing logistic regression, MLP, and CNN models from scratch. Some models achieved 95%+ accuracy and saw training time reduced by up to 50% compared to baseline.



    Contact Information

    Email: haariswaleh@gmail.com
    LinkedIn: https://www.linkedin.com/in/haaris-waleh/
    GitHub: https://github.com/hwaleh



    Resume

    Click here to download my resume.


    to top