Applied Science - Time Series
Core
Design, train, and deploy state-of-the-art time-series foundation models (Timestral) for demanding forecasting problems in industry, from power grids to trading desks.
Role type
Senior Applied Scientist (Time Series)
Builds
Next-generation Timestral models and production-ready AI solutions for industrial and financial clients
Domain
AI/ML, Time Series Forecasting, Energy, Finance
Deliverable
production ML models
Required skills
PyTorch or JAX, deep learning, large language models, time-series foundation models, probabilistic forecasting, multivariate series, Python, model evaluation, data generation, cross-functional collaboration
Preferred skills
PhD or Master in ML/Statistics/Math/Physics, research on time-series foundation models, LLM fine-tuning, agent building, industry experience in energy/telecom/finance, open-source contributions
Technologies
PyTorch, JAX, Python, GPUs, cloud/on-premise environments
Responsibilities
Run pre-training and post-training on GPU clusters; generate and curate large-scale time-series datasets; push the frontier of Timestral architectures and capabilities; fine-tune models on client data; develop tools for data generation and model deployment; collaborate on complex use cases involving agents and LLMs; manage research projects and client communications
Seniority
Senior, hands-on IC
