The Role

We work with sensor data to help our users monitor loved ones and people in care. Our data is fundamentally rooted in time, so it is critical that our data scientists have experience working with time series data. Much of our sensor data is noisy and largely unlabelled, making experience with signal processing techniques especially valuable for cleaning, feature extraction, and building reliable models.

Additionally, we are producing more trends and longer-term views of this data to deliver meaningful insights over time. We’re also looking for someone who can help us develop and validate forecasting models that provide clear, actionable information for our users.

Experience with novel approaches such as transformers or other deep learning techniques would be valuable for tackling complex modeling challenges. Similarly, familiarity with large language models (LLMs) and data-to-text generation would be useful, helping to make complex data outputs more accessible and user-friendly.

Responsibilities

• Analyse and model time series sensor data to identify trends, anomalies, and forecasting opportunities.
• Apply signal processing techniques to clean noisy, largely unlabelled data and extract meaningful features.
• Develop and validate robust forecasting models to support long-term trend analysis and actionable insights for care monitoring.
• Optionally explore novel approaches such as transformer-based or other deep learning models to improve predictive performance.
• Optionally contribute to work with large language models (LLMs) and data-to-text generation to make complex outputs more accessible and user-friendly.

Background

• 3–5 years of experience in a data science or analytical role.
• Experience working with time series data or familiarity with sensor data settings.
• Proficiency in at least one of Python, R, or Julia.
• Strong SQL skills for data extraction and manipulation.
• Experience working with large, noisy, and often unlabelled datasets.
• Nice to have: experience with deep learning approaches (e.g., transformers), large language models (LLMs), or data-to-text generation to enhance modeling and user-facing insights.