Applied Computing

Researcher

• Own Orbital’s foundational time-series modelling stack
• Design, validate, and deploy foundational time-series models under real-world constraints
• Work with customer engineering teams to fine-tune models for their plants
• Own the full lifecycle from theoretical formulation and experimentation to real-time inference, uncertainty estimation, and continuous production retraining
• Design architectures for forecasting, classification/anomaly detection, and optimisation/control-adjacent tasks
• Develop hybrid approaches combining classical statistical models, deep learning, and physics-based constraints
• Integrate conservation laws, process constraints, and differential-equation-based priors
• Design uncertainty-aware models and quantify confidence under sensor drift, failures, regime changes, and sparse or delayed ground truth
• Containerise and deploy models using Docker on AWS/Azure, including EKS, ECS, and SageMaker
• Build or integrate CI/CD pipelines for training, evaluation, rollout, rollback, and automated retraining
• Define back-testing and evaluation protocols and build automated benchmarking pipelines
• Compare classical baselines with modern deep-learning approaches
• Ensure model claims are defensible to customers, partners, and internal stakeholders

Time Series Researcher – Forward Deployed