Founding ML Engineer (Scientific ML/PINNs)
InTensors
- Abu Dhabi
- Permanent
- Full-time
- Architectural design: In addition to standard MLPs, you will develop and deploy models with innovative architectures such as neural operators, graph neural networks, or manifold learning architectures, optimized for scientific data.
- Physics integration: Embedding natural laws into neural networks to ensure realistic results.
- Optimization & scaling: Ensure that complex physics-informed models remain computationally efficient, focusing on memory management and training stability for high-dimensional PDE solvers.
- Validation frameworks: Build rigorous testing pipelines to ensure model outputs remain within the physical feasibility bounds defined by our scientific team.
- Initial phase: Compensation is fully equity based.
- Post funding phase: Upon a successful fundraising (currently in progress), this role transitions to a base salary + equity package.
- PhD in computer science, machine learning, or computational physics is highly preferred. We will also consider candidates with a Master’s degree and a strong track record of professional experience in developing SciML models.
- Expertise in physics-informed neural networks (PINNs), DeepONet, or neural operators.
- Ability to design advanced and optimal architectures that extend beyond the standard MLP architecture to build efficient and scalable models for scientific discovery.
- Strong skills in PyTorch, JAX, TensorFlow, Keras, or ONNX.
- Knowledge of CUDA and GPU acceleration for optimizing custom layers and high performance tensor operations.
- A track record of peer-reviewed publications or a documented history of building and scaling complex SciML models.