← All roles

Modelling Resident

Adaption· Modelling
Apply on company site
RemoteContractPosted 4mo ago

About the role

A modelling residency embedded in production and research projects, mentored by the modelling team. An 8-month remote contract position focused on frontier innovation in efficiency, gradient-free exploration, real-time learning, and interface design.

Responsibilities

  • Innovation: Build the future of adaptive algorithms that continuously learn. Co-design algorithms that react in real-time to product signal and feedback, and explore new ways of capturing feedback that make those algorithms better
  • Cross-Stack Optimization: Collaborate across software, hardware, and algorithmic domains to drive system-wide efficiency gains
  • Measure What Matters: Real-world impact and product signal as the measure of value; ensure algorithms can interact with the world and leverage real-world feedback

Qualifications

  • A degree or equivalent research/engineering experience in a computer science field.
  • Genuine interest in, and at least one project touching: model efficiency, synthetic data, interfaces, real-time alignment, or algorithmic optimization
  • Systems thinking — the ability to understand and optimize across the full ML stack
  • Strong Python skills and experience with deep learning frameworks (PyTorch, JAX, or TensorFlow)
  • Familiarity with model optimization techniques such as RLHF and fine-tuning
  • Strong communication and self-awareness — you know how to collaborate in a remote environment and you're open to feedback
  • You care about technical excellence and last mile impact. We value impact more than effort, and about owning outcomes end to end.

Skills

Required:PythonDeep learning frameworks
Nice to have:Model optimization techniques
Last updated Sep 10, 2026. Interested? Apply on the company site.
Adaption logo

About Adaption

Everything intelligent adapts. So should AI.

Adaptive DataAdaptive IntelligenceAdaptive InterfacesproductAI that adaptsAI that learnson-the-fly malleable datasetscontinual learning

More open roles at Adaption