Christopher Collins, Data Science for Pricing / Logistics / Sports

Location: London, England, United Kingdom

Remote only: Primarily works remotely

Freelancer code of conduct signed

Read the Malt code of conduct

Verified email

16Followers

22Repos

0Gists

Check out languages spoken by Christopher

Agatha Frydrych

Backend Java Software Engineer

Baptiste Duhen

Fullstack developer

Amed Hamou

Senior Lead Developer

Audrey Champion

Web developer

Signup to reveal

Top skills

Additional skills (19)

Check out Christopher's expertise

Digital CV: christophercollins.mydurable.com

LinkedIn: linkedin.com/in/christopher-collins-029731125

Welcome to my profile! This bio should give a quick intro to myself, but I highly encourage you to contact me personally, where I can share my projects and experience in much more detail whilst also getting to know you and your requirements better.

Im a Data Scientist by trade, but spent the majority of my career in startups where I have had to fill a variety of roles. Aside from traditional data science work, I have built data pipelines, created microservices, handled large-scale deployments with AWS and managed cross-functional teams. Being able to do a bit of everything makes me particularly well suited to small businesses trying to achieve great things whilst minimising their headcount.

I currently work as Head of Data Science for Shift, a logistics startup in London, on a part-time basis after running the team full-time for 2.5 years. Most notably, I delivered a data driven pricing framework which achieved a 20% increase in margin and an automated route building engine, dubbed 'Smart Routing', which is able to solve large complex vehicle routing problems. With Smart Routing, we were able to achieve a 10x uplift in bookings (primarily through acquisitions of new businesses) without increasing the size of our operations team.

Before working with small-scale startups I honed my trade at Glovo, a Barcelona based delivery company which was sold in 2023 for $2.3 billion.

When I have free time, I also work on a number of personal projects, including a sports trading model for horse racing and a language app which uses large language models (LLMs) to help users improve their writing skills in other languages.

Key Technologies: Python, AWS, Docker, Github

Proven Industry Experience: E-commerce, Logistics, Sports, Retail

Shift

Head of Data Science

TRANSPORTATION

November 2021 - Today (3 years and 1 month)

Initially working as the only data scientist in a fast paced startup, with the aim of leveraging the company's data to drive growth across the business through a number of high impact projects, including:

Dynamic Pricing - A machine learning approach to generating quote prices for customers. Compared to the original pricing structure, this model was able to improve our margin by 20% with minimal impact in conversion rates.

Smart Routing - An AI approach to route generation for delivery drivers (routes are traditionally generated manually by a team) using optimisation and constraint programming techniques. This project has enabled the company to scale whilst keeping the size of the operations team the same.

Developed a cloud-based data science productionisation framework for fast/secure deployment of models, which can be accessed in a uniform way by developers. This has allowed for seamless integration of models into the delivery app backend.

Recently started to hire more Data Scientists and build a team around developing the projects above, along with future projects in the realms of deep learning and optimisation.

Python MongoDB AWS Optimisation

FSB Technology

Quantitative Developer

TECH

April 2021 - November 2021 (7 months)

London, UK

Developing predictive models for different sports markets, which are used to supply prices to the betting markets sold by the business.

Frequent implementation of Bayesian probability theory, developing models in Python then rebuilding/deploying in Java.

Introduced several new betting markets for football and horse racing.

Huge emphasis on efficient coding practices, as prices can be requested multiple times per second on hundreds of different markets, and need to be shown to thousands of customers with minimal latency. Without a doubt this role improved by technical coding skills.

Python BitBucket Java MongoDB

Nuclio Data School

Teacher

EDUCATION & E-LEARNING

April 2021 - May 2024 (3 years and 1 month)

London, UK

Nuclio is a digital school, offering a range of online data-related courses, whom I work with in addition to my day job. Whilst I find teaching very fulfilling, these courses have also improved my public speaking and ability to communicate complicated concepts. The role includes:

- Teaching 12 hours of data visualization classes to students on Nuclio's Data Science Mastery course.

- Mentoring a team of data science students for their final project.

Check out Christopher's experience