Vetted on a call by a Wave consultant, not scraped. Up to £80,000.
Physics masters turned pricing data scientist, currently on dynamic pricing and time-slot optimisation at a consumer logistics marketplace. The method is the interesting part. Causal inference, debiased machine learning, price elasticity and demand cannibalisation forecasting, plus variational autoencoders and genetic algorithms carried over from the physics side. Python, C++, AWS, Metaflow, PyTorch, CatBoost.
Looking forFeels boxed in by a mid-level remit and a culture that iterates slowly and avoids risk. Wants a startup or scaleup where they own a domain and get to use the harder techniques rather than write about them.
How they came acrossCommunication rated 10 out of 10, and clearly driven by the intellectual side of the work rather than the title. Wants peers who'll push back on the method, and said as much.
Python, SQL, C++ · PyTorch, scikit-learn, CatBoost, pandas · AWS, Metaflow, Weights and Biases · Causal inference, dynamic pricing
4 employers in 3 years, a central bank internship and a space research agency internship, then a vacation rental business and now a consumer logistics marketplace. Average tenure 4 months.
Only a few months into the current job, flagged internally as a possible seniority mismatch for senior briefs, and the record puts them in Manchester while the call notes put them in London.
Masters in Physics, a Russell Group university.
Full stack data scientist at a large payments business with a real engineering bias. Built an evaluation framework for data products shipped as a GitHub Action that 70% of the analytics organisation adopted at a 90% pass rate, and a 12-month commercial forecasting system used by the C-suite that took MAPE from 35% to 26%. Python, SQL, GCP, Vertex AI, XGBoost and PyTorch, and a daily Claude Code user.
Looking forOn paper a product data scientist, in practice doing far more, and there's no role at the current employer that fits the actual remit, which is causing friction over ownership. Wants an AI-first company at Series B or C with autonomy and someone senior enough to learn from. Has ruled out anything over 2,000 people.
How they came acrossCommunication rated 10 out of 10 and operating well past the title. Blocked from promotion by a global tenure policy despite outperforming, and was measured about it rather than resentful, which says something.
Python, SQL, dbt, DuckDB · GCP, BigQuery, Vertex AI, AWS · PyTorch, XGBoost, MLflow · LLM evaluation, RAG
6 employers in 7 years including internships at two very large technology companies, a crypto media business and an insurance consultancy, now a large payments business. Average tenure 9 months.
Under a year into the current job with 6 employers in 7 years behind it, though several were internships, and they've ruled out large corporates and late-stage funding cycles outright.
Masters in Data Science and Machine Learning, and part way through a second masters in machine learning at a Russell Group university, tracking a distinction.
Production machine learning at a gas utility, not notebooks. They architected the data science framework and MLOps strategy, which standardised model deployment across the organisation and cut audit cycle times. Built anomaly detection and probabilistic digital twin models for leak detection that dropped asset failure rates by 10% and found leaks up to 2 weeks earlier. Databricks, PySpark and MLflow day to day, plus computer vision for worksite damage assessment.
Looking forSix years in, and the company is reorganising away from building data and AI capability in-house, which makes their job harder to do properly. Wants somewhere with a clear technical direction where they can keep learning.
How they came acrossMentors 3 junior data scientists and brought it up unprompted. Engaged properly with the roles we put in front of them and reworked the CV for them without being chased.
Python, PyTorch, scikit-learn, XGBoost · Databricks, PySpark, MLflow, SQL · AWS, SageMaker, Bedrock, Docker
3 employers over 9 years, from an NHS commissioning support unit into an agritech weather forecasting business and then 6 years at a gas utility. Average tenure 1 year 9 months, longest stint 2 years 10 months.
Asking £70,000, which is under the market for this depth, so expect a counter-offer and other processes running. Nothing on file that reads as a deal breaker.
Masters in data science from a Russell Group university. TensorFlow developer certified.
Tell me which came closest and I recalibrate, then come back with 10 to 15 more inside 24 hours.
Show me the wider bank