All roles

Research

Member of Technical Staff, Time Series & Predictive Modeling

About the role

Bake AI builds environments, evaluations, and sandbox infrastructure for self-evolving AI. We are looking for a Member of Technical Staff with strong expertise in time series, predictive modeling, or closely related applied machine learning.

You will design, train, and improve models that turn temporal and structured data into reliable predictions. Your work will span problem formulation, data preparation, model development, and rigorous evaluation, with an emphasis on predictive performance, generalization, and practical value.

We are looking for someone who combines strong modeling intuition with careful experimentation and substantial implementation ability. You should be comfortable taking an open problem from raw data to a working model, explaining what drives its performance, and identifying where its predictions become unreliable.

What you’ll do

  • Develop models for forecasting, regression, classification, and other prediction tasks involving temporal or structured data.
  • Investigate and improve model architectures, representations, features, training objectives, and optimization methods.
  • Evaluate statistical methods, gradient boosted trees, neural networks, and pretrained models as appropriate to the problem. Establish strong baselines and understand when additional complexity is justified.
  • Build reproducible data and training pipelines that account for missing observations, irregular sampling, noisy labels, and the availability of information at prediction time.
  • Improve generalization across time periods, datasets, and operating conditions. Investigate distribution shifts, uncertainty estimation, and forecast calibration.
  • Design reliable experiments using appropriate validation splits, rolling backtests, ablation studies, and metrics aligned with the intended use.
  • Connect predictive outputs to downstream decisions and measure their value against clearly defined objectives and constraints.
  • Collaborate with researchers and engineers to integrate modeling workflows into environments where AI agents can inspect data, propose methods, run experiments, and assess results.

What you’ll bring

  • Demonstrated strength in time series forecasting, predictive modeling, or closely related applied machine learning, supported by substantial research or engineering work.
  • Evidence of personally improving predictive performance, generalization, or downstream outcomes. You can explain the problem, your contribution, the evaluation design, and the limitations of the result.
  • Strong foundations in statistics, probability, and machine learning, including regularization, model selection, uncertainty, and the relationship between training objectives and evaluation metrics.
  • Strong Python programming skills and practical experience with scientific computing and machine learning frameworks such as NumPy, pandas, scikit-learn, or PyTorch.
  • The ability to work with imperfect data, diagnose model failures, and distinguish useful patterns from leakage, spurious correlations, or artifacts of the evaluation setup.
  • The judgment to choose suitable methods, challenge assumptions, and carry an ambiguous project through to reproducible results.
  • Clear technical writing and productive collaboration with researchers and engineers.

Preferred experience

  • Multivariate, probabilistic, hierarchical, or long horizon forecasting.
  • Time series foundation models, representation learning, transfer learning, or adaptation across domains.
  • Deep expertise in predictive modeling for tabular data, including feature development, ensemble methods, and learning from limited or noisy labels.
  • Anomaly detection, change point detection, or modeling under distribution shift.
  • Applied modeling in areas such as demand planning, energy, infrastructure, finance, or industrial systems.
  • Automated experimentation, agent driven model development, or benchmark construction.
  • Research publications, open source contributions, or deployed systems with substantial personal responsibility.

How to apply

To apply, email careers@bakelab.ai with your resume and a brief description of a forecasting or predictive modeling project you owned. Describe the problem, the data available at prediction time, the methods and baselines you considered, how you evaluated the results, and what you learned. Links to relevant code, papers, or technical writing are welcome.