Skip to Content

About Me

My name is Stephen (steveoni), I love creating things that are data driven. Currently, I found my self at the frontier of enabling data science and machine learning in the Browser using Javascript (Nodejs, Typescript); by building and contributing to opensource project such as Danfojs, Dnotebook,Datacook e.t.c.

As a machine learning Engineer, i've worked in the areas of computer vision, Natural Language processing and Model Development. And my Research interest is in Privacy AI.

Here are a few technologies I've been working with:

  • JavaScript (ES6+)
  • React
  • Pytorch
  • Tensorflow
  • Node.js
  • Python
  • Docker
  • PyData stack

Where I’ve Worked

Machine Learning Engineer/ Researcher @ Data Science Nigeria (DSN)

Feb 2020 - Oct 2020

  • Participated in the goal of raising 1 million AI talent in Nigeria in the next 10 years by teaching different classes comprising of (undergraduate, graduate and working class) in machine learning
  • Research and Build a brand sentiment analysis for banks tackling code switiching (mixing of pidgin with english) in customers text . (A team of 2)
  • Spear head the research in building fast customer response bot for financial instuite using Natural language processing to tackle lags in customer's response and also managing attention.
  • This research was also presented as a poster in 2020 Neurips. Also the bot was deployed on aws and the tools use are Pytorch, boto3, twitter api, fastai.
  • Did research on how to eliminate hunger in Nigeria and we propose the use of NewsVendor model to prevent food wastage. the paper was submitted as a poster in Global EC20 Conference.
  • Spear-head the development of cv ranker (document retrival ) to assist HR in filtering out cvs.

Some Things I’ve Built

Other Noteworthy Projects

view the archive
  • Data-Art

    A no-code platform to enable data scientist handle data like an artist.

    • Javascript
    • Danfo.js
    • React.js
    • Chart.js
    • React-Table
  • DnoteBook React

    Re-build Dnotebook (A jupyter notebook in javascript) with React.js

    • Javascript
    • Danfo.js
    • React.js
  • Exploring Language Models

    The article is a set of series exploring language models from rnn to transformers and from transformers to GPT2 to using HuggingFace

    • Python
    • Pytorch
  • Facial Recognition with Javascript

    The article shows how to perform object detection on images and video in the browser.

    • Javascript
    • Faceapi.js
    • Tensorflow.js
  • Deploy your language model on google cloud

    A project driven article on how to deploy a Language model on google cloud

    • Flask
    • Google cloud
    • Docker
    • Pytorch
  • Libsvm-wasm

    An opensource contribution; to port libsvm (c and c++) to Typescript.

    • Typescript
    • Wasm
    • C
    • C++

What’s Next?

Get In Touch

I will be glad to contribute to any opensource project regarding Data science/Machine learning