Better decisions through causal data science
I help data teams, marketing analysts, and organisations move beyond correlation‑based thinking through expert consulting and practical workshops in causal inference, Bayesian analytics, and marketing measurement. Based in Scotland, working internationally.
How I work
Two ways to engage, depending on whether you need answers or capability.
Upskill your team
Practical, expert-led training in causal inference, marketing measurement, and experiment design. Delivered remotely or in-person, tailored to your team's level and domain.
See workshop offerings →Expert analysis for your team
Project-based or embedded engagements. I help teams quantify causal impact, design experiments, and build decision-ready Bayesian models. Delivered as focused sprints, model reviews, or ongoing advisory.
For larger projects, I can bring in additional expertise through PyMC Labs.
Selected work
Representative write-ups from client collaborations, with more in my technical portfolio.
Improving marketing effectiveness with Bayesian media mix modeling and time-varying effects.
Impact: faster iteration on channel allocation decisions.
Read the write-upBayesian A/B testing at scale, with speedups that keep decision cycles fast.
Impact: reduced analysis time from days to hours.
Read the write-upDetermining whether promotions drive incremental sales or cannibalize demand.
Impact: clarified where to invest vs. where to reduce spend.
Read the write-upDeeper case studies and technical write-ups, including work with the Bill & Melinda Gates Foundation and Colgate-Palmolive.
Impact: applied Bayesian and causal methods across public health and consumer research.
Visit the portfolioClient work includes collaborations with PyMC Labs.
Open-source leadership
Building tools that make Bayesian causal inference practical for real-world teams.
Lead developer
CausalPy is an open-source package for modern causal inference workflows, with a focus on quasi-experiments when randomized trials are not feasible. It has 1,000+ GitHub stars and 22k monthly downloads.
View on GitHubContributor
Contribute major features and clear documentation to help teams maximize leverage from Bayesian media mix models in pymc-marketing.
Repository · My PRs
About Dr Benjamin Vincent
I run InferenceWorks Ltd as a one-person consultancy. I earned a BSc and DPhil from the University of Sussex (UK) and specialize in Bayesian statistics and causal inference, especially quasi-experiments when randomized trials are not feasible.
Working style
Clear explanations, practical outcomes, and rigorous understanding. I deliver transparent analysis, code, and guidance your team can use.
Credentials
BSc and DPhil (PhD), University of Sussex, UK. 15 years in academia and in industry since 2021.
Peer‑reviewed publicationsLet’s talk
Tell me about the problem you're working on. I'll respond within two business days.
NDA‑ready. Rates shared on request.