Description
Imagine negotiating a major steel, energy, agricultural, or industrial-material contract without knowing whether the underlying market is likely to rise or fall six months from now.
Monq wants to turn that uncertainty into a modelling problem.
The Senior Data Scientist / Researcher will build the price-intelligence layer behind the company's procurement platform, developing forecasting systems that help enterprise buyers understand what a contract should cost before negotiations begin.
The modelling work goes well beyond one standard time-series algorithm. Potential approaches include gradient-boosted ensembles, Gaussian processes, recurrent neural networks, neural state-space models, Bayesian techniques, and hybrid symbolic/statistical architectures.
Alternative data plays an important role. Forecasting signals may come from satellite imagery, shipping activity, weather, procurement indexes, commodity-market data, and news sentiment rather than only historical price series.
This creates a substantial feature-engineering challenge. Signals arrive at different frequencies, carry different levels of noise, and may have changing relationships with the target variable as markets move through different regimes.
The scientist will own the modelling lifecycle from hypothesis and feature development through model selection, validation, uncertainty quantification, deployment, and monitoring. Time-series validation needs particular care because ordinary random train/test splits can produce misleading results when applied to market data.
Probabilistic output also has to become useful product information. A procurement professional needs to understand not just that a model predicts a certain price, but how uncertain that forecast is and whether the expected movement is strong enough to affect a real negotiation strategy.
Useful experience includes XGBoost, LightGBM or CatBoost; LSTM, GRU, TFT or N-BEATS-style architectures; PyMC or NumPyro for probabilistic modelling; MLflow or Weights & Biases for experiment management; and explainability techniques such as SHAP.
Monq is seeking a senior practitioner with substantial applied data-science or quantitative-research experience, particularly around forecasting, time series, commodity markets, energy, finance, or other domains where predictions have to survive noisy and rapidly changing real-world conditions.
Unlike a purely academic research role, the objective is to get models into production. The scientist will work closely with engineering so successful research can become part of a live AI product used during enterprise negotiations.
Curated opportunity. Please verify details and apply via the original link below. No Signals are required for this project/job.
https://careers.monq.io/jobs/7937751-senior-data-scientist-researcher