Victor Baker


Data engineer with a passion for problem solving and a lifelong interest in Artificial Intelligence. I am proficient with Python, SQL, JavaScript, and HTML/CSS. I have experience using Python to implement machine learning solutions with neural networks, and exposure to computer vision, natural language processing, image classification, and time series analysis.

Skills

Influencing Decisions

  • Relating technical concepts
  • Effective communication
  • Patience, consistency
  • Build trust
Work Experience

Data Engineering

  • Python, pandas, numpy, SQLalchemy
  • MongoDB, SQL, Postgres
  • MATLAB, Excel VBA
  • Git, flask, aiortc, aiohttp
Projects
FullColorPrimary Icon

Machine Learning

  • Tensorflow, keras, colab
  • Neural networks, deep learning
  • Computer Vision, openCV
  • Scikit-learn, numpy, pandas
Projects

Frontend Development

  • Javascript
  • HTML
  • CSS
  • plotly.js, d3.js, leaflet.js
Projects

Data Visualization

  • Tableau
  • matplotlib, pandas
  • plotly.js, d3.js
  • leaflet.js, mapbox, googlemaps
Projects

Customer Service

  • Customer Service Lead
  • Retail Management
  • Positive attitude
  • Follow through
Work Experience

Facial Emotion Recognition

This app receives a video stream from the users browser, and returns an emotion prediction based on facial expression.

Face detection is implented using a Haar Cascade. The portion of the image containing a face is passed to a Convolutional Neural Network for emotion prediction.

Tools Used: Python, Javascript, Tensorflow, openCV, webRTC, aiohttp, aiortc, and others
Explore the app Check out the code

Citibike Data Pipeline

This project extracts monthly trip data from amazon S3 buckets, and loads them into a SQL database. The 112 million trips are aggregated to create a Tableau story on bicycle longevity.

Data extraction and transformation is done in memory, with the clean data then appended to a SQL table on disk.

Tools Used: Python, SQL, Tableau, pandas, requests, io, zipfile, SQLalchemy.
Explore the app Check out the code

Kepler Planet Classification

This project tests the efficacy of machine learning algorithms at predicting the classification of planetary candidates. Ease of implementation and final accuracy are compared for Random Forest and simple deep learning models.

The Keras Tuner API is used to execute Hyperband hyperparameter tuning on the neural networks.

Tools Used: Python, pandas, scikit-learn, Keras, and TensorFlow
See the analysis Check out the code

Bacterial Colonization

This interactive plotly dashboard visualizes bacterial oligarch species inhabiting belly buttons. The user is able to view study data for individual participants

Javascript and D3 are used to parse through the JSON file, bind data, and dynamically write html to the page.

Tools Used: Javascript, d3,js, and plotly.js

Explore the app Check out the code

Shark Sighting API

This app explores site preferences for sharks using geoJSON data and interactive maps. The goal was to test the hypothesis that tiger sharks display site fidelity for shipwrecks.

The raw data from the API is cleaned, and then added to a leaflet map as a custom icon for each shark species or ship.

Tools Used: Python, JavaScript, SQL, HTML/CSS, pandas, SQLalchemy, textillate.js, bootstrap, d3.js, and leaflet.js
Explore the app Check out the code

About Me


I enjoy reading, hiking through local parks, playing video games, and chocolate ice cream.