Machine Learning & Simulink Engineer

About us

We are a company based in London and Turin building physics-grounded models and simulation infrastructure for Embodied Intelligence. We help industrial, automotive and robotics teams train, test, evaluate, and debug intelligent systems before deployment, with an emphasis on observability, verification, and scalable simulation.

Our work sits at the intersection of machine learning, physics-based simulation, control, and production-grade engineering. The team combines experience across reinforcement learning, robotics, physical modeling, large-scale AI systems, and enterprise deployments.

We are small, technical, and execution-driven, working directly on customer-critical systems where physical fidelity, reproducibility, and safety constraints matter.

Compensation & Collaboration

You will work on a short-term automotive project for a customer. This is a temporary position, but we are happy to discuss a longer engagement at the end of the project if there is a strong fit with the company's needs and we onboard more customers in the same project area.

Option A

• contractor work at €25-40/hour, with payment structured as 25% at project start, 25% at mid-project, and 50% upon delivery
• Exact project scope and time commitment will be discussed and agreed during interview and onboarding.
• We can support for contractor work with a platform to manage contractor work if you have never done this before but you are interested in the position.

Option B

• €40-50k RAL equivalent, for employment contract lenght between 1-3 months, renewable if customer project is successful
• Exact project scope and contract lenght will be discussed and agreed during interview and onboarding.

Interview

• First technical interview with the CTO
• Second interview with the CEO
• Interviews will be scheduled via google meet, we can accomodate interviews during evenings and weekends.

Responsibilities

• Understand, debug, and refactor existing MATLAB/Simulink and Simscape thermal-system models, including heat-pump, coolant-loop, battery, cabin, motor, and power-electronics dynamics.
• Resolve simulator bottlenecks related to solver stability, speed changes, algebraic or numerical failures, state initialization, and long warm-up periods.
• Expose controllable initial conditions, state variables, observations, and outputs through robust Simulink buses and programmatic interfaces.
• Convert the simulator into a headless, scriptable environment that can run batch simulations, parameter sweeps, and parallel campaigns without GUI dependency.
• Build and maintain the Python-Simulink integration layer for reset, step, action, observation, logging, and experiment orchestration.
• Support the development of an MPC or optimization-based oracle and use it to generate expert trajectories and training targets.
• Support ML components in Python, including predictive models, neural surrogate models, imitation-learning pipelines, and reinforcement-learning environments.
• Generate representative mission profiles and synthetic datasets covering urban, extra-urban, payload, climate, route, and edge-case conditions.
• Design comparative tests between deterministic controllers, optimization-based strategies, and AI policies.
• Implement safety constraints, deterministic fallback logic, and validation checks before real-data and vehicle testing.

Qualifications

• The candidate has held the job title Machine Learning & Simulation Engineer, Python/MATLAB-Simulink Engineer, or its equivalent.
• Must have experience with MATLAB and Simulink, ideally including real model debugging or simulation project work.
• Requires Python engineering skills with NumPy, SciPy, pandas, and preferably PyTorch or another ML framework.
• Must have an understanding of dynamic systems, state-space modeling, control systems, numerical simulation, or physical systems modeling.
• Ability to work with time-series data, experiments, logs, and reproducible scripts.
• Junior to mid-level seniority with strong practical fundamentals, fast learning ability, and hands-on experience.
• Experience with Simscape.
• Experience with solver configuration, stiff ODE/DAE systems, zero crossings, discontinuities, algebraic loops, and model initialization.
• Experience with Simulink Bus, Model Reference, Data Dictionary, parameterization, or programmatic simulation.
• Experience with automotive thermal management, HVAC, batteries, BMS, electric powertrain, or energy-management systems.
• Available to start immediately or within the next 1 to 2 weeks.
• Full remote position.
• Fluency in Italian or English.
• Experience with headless Simulink execution, batch simulation, profiling, and parallelization.Background in MPC, optimal control, design of experiments, numerical optimization, system identification, surrogate modeling, imitation learning, or reinforcement learning.
• Availability to use Slack as a communication channel.
• Proactive and high-touch work communication style.