Home Intelligence Engineer
What if a room could understand the moment?
MVP → real-world validation · Expression of interest
Help turn an intention such as “I am going to watch a movie” into an environment with the right light, comfort and privacy. The hard part is the translation from human language to an explicit target state that a physical system can safely deliver.
What you would own.
- Build behaviour models and scene synthesis against a shared home semantic graph.
- Separate language interpretation from deterministic capability checks and execution.
- Make resident corrections useful to a preference-learning loop without hiding why a decision was made.
What you bring.
Applied ML, personalisation, control systems or decision-making under uncertainty. You care about explainability at the moment a system acts.
A meaningful first contribution.
Take a defined scene-synthesis path from design to a repeatable evaluation against realistic home scenarios, including ambiguous requests and resident corrections.
The environment around you.
Work across product, spatial models, software and real operating constraints. We value evidence over theatre, clear ownership over hand-offs and a willingness to make the difficult parts understandable. Agent-assisted development is part of that approach; responsibility remains with people.
Make this your next chapter.
Introduce yourself and share a project, technical note or example of a difficult problem you have solved. Include your location and the kind of contribution you want to make. Role scope, location, timing and employment terms will be discussed directly.
Express your interest →