George Willis / Data scientist

About

George Willis

I’m a data scientist in London, working on internet and mobile-network performance. Much of the work involves making sense of messy measurements: investigating unusual patterns, checking whether comparisons are fair and explaining what the data does and doesn’t show.

Before Cisco, I worked at SamKnows. Across both, I’ve worked on customer investigations, data pipelines, regulatory reporting and tools that make analysis easier for other people to use. I enjoy the combination of working through a technical problem and explaining the result to someone who needs to make a decision.

My research background is in spatial analytics, networks and urban change. I’ve worked with taxi trips, links between websites and satellite imagery, often using data collected for one purpose to investigate something else. My published work includes research on Uber and New York taxis, and using web links to predict regional trade.

Outside work, I tend to build things I’d like to use. Giggity grew out of missing the more niche London gigs: even with discovery apps and ticket sites, I still found myself checking venue programmes separately and trying to keep track of changes. TransitFlow is a more playful experiment: could I build a metro-network model, then apply the same approach to another city?

These projects also give me a reason to learn beyond my usual data-science work. I’m comfortable with the analysis, but a lot of the application engineering has been new territory. I’ve enjoyed figuring out how to connect the data, interface and deployment, rather than stopping at a notebook.

An earlier version of Giggity got a little over-engineered, but gave me hands-on experience with FastAPI, PostgreSQL, SQLAlchemy and Alembic migrations, alongside a containerised local database using Podman. I also prototyped vector-based event matching with pgvector. Rebuilding around a simpler collection and publication process became part of the learning: understanding the tools, then deciding which ones the app needed. More on those choices →

This site brings together those projects, published research and a few older analyses. Some are finished pieces of work; others are still giving me reasons to keep learning.

Current role
Data Scientist, Cisco ThousandEyes
Main interests
Data analysis, spatial networks and useful AI tools
Languages
R, Python and SQL
Research
PhD - University of Birmingham. Thesis submitted in 2023.
Contact
gcw519email@proton.me
LinkedIn
uk.linkedin.com/in/gwilisli26 ↗

Experience & research

Cisco ThousandEyes Data Scientist · London

Internet and mobile-network performance analysis, customer investigations, data pipelines, regulatory reporting and internal analytical tools.

SamKnows Data Scientist · London

Broadband and mobile measurement pipelines, customer investigations, regulatory reporting and tools supporting wider analysis.

University of Birmingham PhD Research in Spatial Networks

Spatial analytics, urban change and network modelling across mobile O/D flows, web hyperlinks and satellite imagery (thesis submitted in 2023).