Vetted on a call by a Wave consultant, not scraped. £110,000 to £120,000.
Owns quote-level dynamic pricing and search pricing at a consumer logistics marketplace, end to end. The causal toolkit is the real draw. A/B testing, CUPED, difference-in-differences, propensity-score matching, uplift modelling and synthetic-control counterfactuals, alongside churn, LTV and BERT classification work. Python, SQL, Snowflake, dbt, Airflow and Databricks underneath it.
Looking forMoving because the data at their current employer runs out before the models get interesting, and they want a richer problem with real end-to-end ownership. Earlier-stage company with equity upside, and 4 days a week in an office is the absolute ceiling.
How they came acrossPresents to the CFO and senior leadership weekly and it shows in how they talk about the work. Rated 9 out of 10 on communication when we vetted them, and every claim came attached to a number.
Python, R, SQL · Snowflake, BigQuery, Databricks, dbt, Airflow · scikit-learn, Transformers · AWS, GCP
3 employers in 6 years, a year into a consumer logistics marketplace after two technology consultancies, one in sports ticketing and one in automotive. Average tenure 1 year 6 months.
Won't do more than 4 days a week in an office, and notice is a month.
Masters in Statistics.
Combinatorial optimisation is the specialism, vehicle routing in particular, and they've been at it about 3 years inside a large online grocery and logistics operation. Writes Python and Scala, builds with hybrid genetic search and metaheuristic frameworks, and carries it through CI/CD to deployment rather than handing over a notebook. Also does tree-based and time series forecasting for hourly and geographical demand.
Looking forWants out of vehicle routing before it becomes the whole CV, and wants a principal-level optimisation person to learn from, which they haven't got now. Fully remote, around 100 people, and 110 is the number they gave us.
How they came acrossCommunication rated 10 out of 10. Built their own interactive portfolio site with Claude Code and a skills-first CV, which tells you how they approach a problem when nobody asked them to.
Python, Scala, SQL · Metaheuristics, hybrid genetic search, Monte Carlo · Time series forecasting, MLOps · GCP, CI/CD
7 employers across 12 years, from an astronomy institute internship and an optimisation software startup through a large retail bank, now a large online grocery and logistics business. Average tenure 1 year 5 months, longest stint 4 years 7 months.
Redundancies at their current employer are the trigger rather than the job itself, and our own recruiter thought the 110 asking was punchy for the level.
Masters in Operational Research with Data Science, a Russell Group university.
Data scientist inside a European venture fund, working across timeseries forecasting, cluster analysis and LLM and RAG applications. Python and SQL as the base, with xgboost, scikit-learn, TensorFlow and Keras, and DBT and prefect on the pipeline side. Came into it from software engineering, including a spell in high performance computing at a motorsport engineering business.
Looking forNothing reliable on file. The call notes attached to this record describe a different person in a different job, so treat anything about motivation or notice as unconfirmed until someone speaks to them.
Python, SQL · scikit-learn, xgboost, TensorFlow, Keras, pandas · DBT, prefect · GCP, BigQuery
3 employers in 6 years, a motorsport engineering business, then a consumer technology company of more than 150,000 people, now a European venture fund. Average tenure 1 year 4 months, longest stint 3 years 5 months.
Prefers in-person working, availability isn't confirmed, and the call notes on file are unreliable, so a screening call is worth doing before any client process.
Masters in Computer Science, a Russell Group university.
Tell me which came closest and I recalibrate, then come back with 10 to 15 more inside 24 hours.
Show me the wider bank