Physicist

Huberta; Sheffield; Remote; GBP 80,000 - 100,000; Full time

Job summary

An early-stage AI startup is seeking a physicist to develop core components of a proprietary AI model. The role involves mathematical formulation, algorithm optimisation, and is fully remote. Candidates should hold a PhD in Physics and have experience applying physics in industry. This position offers significant ownership and the opportunity to lead teams as the company scales, with competitive compensation and equity.

Benefits

Equity as an early team member

Flexible remote work

Opportunities for growth and leadership

Qualifications

• 3+ years of industry or research experience applying physics.
• Experience in theoretical, statistical, or computational physics.

Responsibilities

• Develop core components of an AI model.
• Design novel learning approaches informed by physics.
• Optimise algorithms for production systems.

Skills

Quantum Mechanics

Multi-Dimensional Tensor Networks

Mathematical formulation

Advanced linear algebra

Information theory

Education

PhD in Physics

Job description

Fully Remote (Europe & UK) | Early-Stage Startup | Stealth Mode

What We're Building

We're a well-funded startup developing a proprietary AI model that takes a fundamentally different approach to machine learning. We're not iterating on existing architectures — we're building something new from the ground up.

We need a physicist who can turn theoretical insights into production systems. We need people who understand the theory and formulate the underlying mathematics before any code is written.

The Role

You'll be developing core components of our AI model, engaging in hands-on work at the intersection of physics, mathematics, and ML engineering. This isn't purely theoretical research. You'll own significant pieces of our model architecture from mathematical foundations through implementation and optimisation.

Small, high-calibre team with real ownership.

As we scale, opportunities to build and lead teams. We're hiring immediately.

What You'll Actually Do

Derive mathematical formulations of abstract concepts. Design and implement novel learning approaches informed by physics and mathematics. Take concepts from statistical mechanics, information theory, and dynamic systems to construct first principles algorithms. Optimise using advanced mathematical structures, such as tensor networks, and techniques like reinforcement learning - make these approaches work in a production system.

Debug why your theoretically sound approach breaks at scale. Fix it. Ship it.

Daily reality includes mathematical derivations and performance optimisation. You'll need to be comfortable moving between theory and systems-level engineering within the same afternoon.

Required Qualifications

Education & Experience :

• PhD in Physics : Theoretical / Statistical / Computational / Applied / Mathematical / Quantum / etc.
• 3+ years of post-PhD experience applying physics in industry or research settings. Non-academic research experience is also accepted.

Core Technical Areas

Physics :

• Quantum Mechanics ( Must have )
• Multi-Dimensional Tensor Networks ( Must have )
• Theoretical physics and statistical physics
• Dynamic systems (energy landscapes, emergent behaviours)
• Modelling and simulation

Mathematics :

• Advanced linear algebra, optimisation, numerical methods
• Information theory
• Probability, statistics
• Graph theory

Highly Valued

• Machine Learning Experience : Specifically in parameter-free learning approaches, Bayesian methods and belief networks, reinforcement learning and graph neural networks, computational optimisation at scale, algorithms and data structures.
• Quantum Information Theory
• Experience working with stochastic data

Who Thrives Here

You work fast in loosely defined environments. Competing priorities don't slow you down.

You own problems end-to-end. If you need to learn something to solve it, you learn it.

You're comfortable with ambiguity and rapid context switching. Startup pace doesn't rattle you.

Clear communicator in English. Self-sufficient but collaborates well.

This Won't Fit If :

• You need complete specs before starting
• You think rigorous means slow
• Current ML paradigms satisfy you
• Proving theorems appeals more than deploying systems

What We Offer

Direct Impact : Build proprietary AI architecture from scratch

Equity : Meaningful stake as an early team member

Growth : Lead teams as we scale

Resources : Equipment, conference attendance, publication opportunities

Autonomy : Real ownership of technical decisions

Flexibility : Fully remote within Europe / UK (Central European timezone business hours)

Compensation : Competitive salary and equity package (discussed during interviews)

Location

Fully remote, Europe / UK-based, available during Central European timezone business hours. May transition to hybrid later.

Apply

Send your CV to [e-mail]

We review applications on a rolling basis and respond promptly to strong candidates.