Theo Arnold-Forster, Machine Learning Developer in London, United Kingdom
Theo Arnold-Forster

Machine Learning Developer in London, United Kingdom

Member since November 4, 2021
Theo is a full-stack data scientist with a solid academic background and four years of professional experience. While working on customer data, he became an expert in improving customer experiences, maximizing revenues, and driving marketing efforts by owning end-to-end projects. With his exceptional skills in analysis, machine learning, and data engineering, Theo enjoys working with challenging systems and strives to use the right tool for a particular job.
Theo is now available for hire

Portfolio

Experience

Location

London, United Kingdom

Availability

Part-time

Preferred Environment

PyCharm, Jupyter Notebook, Python, SQL, Metabase, R

The most amazing...

...thing I've built is a model that was deployed in a call center, massively driving sales and helping staff beat their targets.

Employment

  • Lead Data Scientist

    2019 - PRESENT
    Muzmatch
    • Built and maintained a data warehouse in Redshift, an ETL system in Python and Airflow, and a BI system in Metabase to analyze digital marketing APIs and multiple production databases.
    • Ran the A/B and multivariate tests for product optimization. Developed and deployed the machine learning models to provide product personalization to millions of customers.
    • Built a small and efficient data team that oversaw all data science functionalities in the company.
    Technologies: Python, Data Science, Machine Learning, SQL, Apache Airflow, Linux, Business Intelligence (BI), ETL, Data Warehousing
  • Data Scientist

    2018 - 2019
    Domestic & General
    • Created a model to predict and prevent complaints from claimants, which significantly reduced the complaints. Led the end-to-end project.
    • Built and maintained a model to order leads at a large-scale contact center telephone campaign that led to significant uplifts in revenue. Managed the project from the beginning to the end.
    • Taught SQL and R to analysts from across the business.
    Technologies: R, SQL, Analysis

Experience

  • Personalized Pricing Model

    The personalized pricing model is a system that delivers personalized product discounts and prices to millions of customers.

    I built the system and led the project from conception to testing, deployment, and evaluation. Furthermore, my responsibilities involved designing a series of randomized tests to collect relevant data for modeling the relationship between price and demand and allocating the optimal expenses to maximize revenue before building the model.

    We deployed the model via a batch process that preprocessed the data for modeling in Redshift and Python before sending updated prices daily to a live production application. The model was evaluated using data stored in the Redshift cluster with a BI system built in the Metabase. It resulted in several iterations of uplifts and significantly increased revenue against repeated control groups.

Skills

  • Paradigms

    Data Science, Business Intelligence (BI), ETL
  • Languages

    Python, R, SQL
  • Platforms

    Jupyter Notebook, Linux
  • Other

    Statistics, Test Design, Machine Learning, Analysis, Metabase, Data Warehousing
  • Tools

    Apache Airflow, PyCharm
  • Storage

    Redshift

Education

  • Master's Degree in Big Data and Business Analytics
    2016 - 2017
    IE University - Madrid, Spain
  • Bachelor's Degree in Biological Anthropology
    2013 - 2016
    University of Cambridge - Cambridge, United Kingdom

To view more profiles

Join Toptal
Share it with others