• Design and deploy AI models for prediction and decision-making on domain-specific operational data
• Advance and extend scheduling and resource optimisation, including multi-objective optimisation, constraint handling, and stable re-planning
• Build end-to-end models that learn from real operational data and improve planning accuracy over time
• Own the full AI lifecycle: data analysis, feature engineering, model development, evaluation, and monitoringTranslate business problems into formal models and measurable outcomes

Examples of What You’ll Build

• Capacity planning algorithms that account for skills, time buffers, cool-down periods, and space constraints
• Intelligent job-to-talent matching with dynamic weighting for in-progress vs. new projects
• Automated project creation from structured and unstructured operational inputs (e.g. PDFs, free text)

Requirements

• Master’s or PhD in Computer Science, Mathematics, Physics or a related field
• Several years of experience in machine learning, particularly with tabular data and time series
• Strong knowledge of at least one deep learning framework (PyTorch is preferred)
• Experience designing optimisation algorithms and heuristics for complex, constrained problems
• Proficient in Python and the data science ecosystem (e.g. pandas, scikit-learn)
• Experience with ML experiment tracking and model deployment (e.g. MLflow, Docker, Kubernetes)
• A self-driven, solution-oriented mindset: you don’t just build models, you understand the business problem behind them
• Fluent in English

Nice to Have

Beyond our core artificial intelligence, we are exploring additional AI capabilities that will flow into the product over time. Experience in this area is not required, but a plus:

• Experience with LLMs or agentic systems (e.g. RAG, information extraction from unstructured data, API-based workflows)

Benefits

• Creative freedom in a young, technically ambitious team
• Direct impact of your work on the product and our customers
• A modern tech stack and a culture that encourages experimentation
• A real-world, domain-rich problem space, not another generic SaaS
• End-to-end ownership of AI features, from data exploration to production deployment