Lead Data Scientist – Demand Forecasting

How do you make our customers happy?

By turning trillions of behavioral, operational and external data points into accurate, probabilistic demand forecasts. Your signal keeps 52,000 partner-sellers in stock, frees working capital, reduces waste and ultimately ensures our 13 million Dutch and Belgian customers always find the products they want, when they want them.

The biggest challenge

Forecasting demand for ~42 million SKUs in a two-sided marketplace where trends, promotions and external shocks change customer behaviour daily. You’ll set the technical vision—and build the models—that turns this complexity into clear, trusted guidance for planners, buyers and automated supply-chain services.

What you’ll do as Lead Data Scientist – Demand Forecasting

• Set technical direction for the entire forecasting stack—from model architecture and feature engineering to MLOps standards, monitoring and alerting.
• Design, ship and own cutting-edge models (Temporal Fusion Transformers, diffusion forecasters, multimodal foundation models) that ingest behavioural, operational and exogenous signals.
• Mentor & coach data scientists and data engineers, raising the bar on modelling quality, code craft and experiment design.
• Drive adoption of probabilistic forecasts and decision frameworks in supply chain and buying.
• Quantify impact through online experiments and KPI uplifts (availability, lost-sales rate, working-capital turns).
• Represent Demand Forecasting in architecture guilds and external meet-ups, sharing

Why you can make a difference

• 6+ years building and operating ML systems in production, preferably at e-commerce or supply-chain scale.
• Proven leadership in setting technical vision, influencing product roadmaps and coaching peers.
• Expert Python (Pandas, PyTorch/TensorFlow, scikit-learn, statsmodels) plus strong SQL/BigQuery; solid software-engineering practices (testing, CI/CD, code review).
• Hands-on with GCP (Vertex AI, BigQuery ML, Dataflow) and workflow orchestrators (Airflow, Dagster) is a bonus.
• Deep knowledge of statistical inference, causal analysis, uncertainty quantification and experiment design.
• Storytelling skills that translate probabilistic insights into clear, actionable narratives for senior (non-technical) stakeholders.

- . You lose interest after the POC and delegate productionisation to “others”.

- . You believe “deeper nets” always beat a well-designed baseline.

- . You prefer working in isolation over co-creating with engineers, planners and product leads.

+ . You love growing people as much as models, and measure success in adoption & business value.

+ . You balance technical depth with pragmatic delivery and crisp communication.

+ . You thrive on autonomy, take ownership of outcomes and welcome accountability.

This is where you’ll work

bol is the leading retail-tech platform in the Netherlands and Belgium. Since 1999 we’ve experimented, learned and scaled—always putting customers first. You’ll join ~3,000 colleagues who believe that better models create better experiences. In our Demand Forecasting team you’ll collaborate daily with Product Managers, ML Engineers, Data Engineers and fellow Data Scientists to deliver the most trusted demand signal in Benelux e-commerce. We celebrate ownership, experimentation and inclusive collaboration—and we measure success by the value we unlock for customers, partners and the planet.

Ready to forecast the future? Let’s talk!