Skip to content
View kobinabrandon's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report kobinabrandon

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kobinabrandon/README.md

Who am I?

I'm a Machine Learning Engineer (MLE) of Ghanaian, Sierra Leonean and British heritage.

After studying Mathematics for much of my early adult life, I now build end-to-end machine learning systems (preferably trained on living datasets) that solve problems I care about. I genuinely enjoy programming, take immense pride in whatever I'm able to create, and am constantly looking for ways to expand my skills.

How I Work

I impose certain professional standards upon myself.

  • I commit to produce code that meets the highest standards of quality that my skills will allow at any given time, and to adopt standard software engineering best practices.

  • Whenever possible, I favour developing end-to-end machine learning applications. This means that I almost always opt for writing modular code in scripts as opposed to Jupyter notebooks. As a result, I try to keep my use of Jupyter notebooks to an absolute minimum, and only use them for EDA (exploratory data analysis), and to quickly tests of pieces of code.

  • I don't use AI coding assistants. In their current and most popular configurations, these tools often suggest code before one has even had the opportunity to think independently, repeatedly encouraging us to choose their suggestions over our own thoughts. I fear that over time, this pattern of behaviour may cause our skills to atrophy, creating dependency and feelings of discomfort in habitual users who are suddenly forced to do without them.

  • I only use LLMs when I have specific questions about how functions within a given library work, or about language features.

Other Interests

  • Beyond my personal professional goals, I would like to somehow apply my skills to contribute in some small way to the technological and economic advancement of my fellow Africans.
  • I enjoy deeply programming, and I have an interest in other aspects of software engineering, though machine learning engineering remains my focus.
  • I also harbour an interest in embedded machine learning, and would like to one day explore that aspect of the field.
  • I would like to add the Rust and Go to my language base. I like Rust because of its complex type system, friendly compiler, and the way it forces you to write code in certain ways. I like Go because it's as friendly as Python, but is statically typed and much faster.
  • I'm learning and enjoying Neovim, the chad editor, and I invite you to try it.

🛠️ Languages and Tools

Neovim  Python  Rust  Pytorch  scikit-learn  Optuna  Comet  Hopsworks  Linux 

Pinned Loading

  1. Hourly-Divvy-Trip-Predictor Hourly-Divvy-Trip-Predictor Public

    An end-to-end batch machine learning system that produces hourly predictions of the number of arrivals and departures that will take place at various stations in Lyft's bike sharing system in Chicago.

    Python 8 2