Physics-Informed ML Engineer for Health Modeling (Remote)

Svitla Systems, Inc.; Argentina; Presencial; ARS 1.800.000 - 3.000.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Remote work

Bonuses for referrals

Vacation days

Learning program

Tech webinars

Descripción de la vacante

Svitla Systems, Inc. is seeking a Physics-Informed Machine Learning Engineer for a full-time role in Argentina. You will design and train PINN-style models for degradation and health prediction, integrating physics-informed losses with data-driven learning in a Python stack.

You will build end-to-end data pipelines, calibrate models with calibration strategies, and collaborate with scientists to operationalize research into production code across US/EU projects.

Formación

• Experience building physics-informed models with a physics-based loss in a real project.
• Strong time-series/sequence modeling on sensor/telemetry data.
• Ability to calibrate mechanistic model parameters to noisy data.
• Expertise in Python data stack and end-to-end data pipelines.
• Ability to read and reason about physics/reliability equations.

Responsabilidades

• Build and train a physics-informed sequence model for degradation and health prediction.
• Design a fusion layer combining physics features with data-driven signals.
• Calibrate stress-proxy parameters and run calibration strategies.
• Harden the Python/Pandas feature pipeline with audits and data-quality gates.
• Document analyses and defend modeling choices to technical stakeholders.

Conocimientos

PINN experience

Time-series modeling

Bayesian calibration

Python scientific stack

Data pipeline

Physics-based loss

Herramientas

PyTorch

NumPy

Pandas

scikit-learn

JAX

Descripción del empleo

Svitla Systems, Inc. is seeking a Physics-Informed Machine Learning Engineer for a full-time role in Argentina. You will design and train PINN-style models for degradation and health prediction, integrating physics-informed losses with data-driven learning in a Python stack.

You will build end-to-end data pipelines, calibrate models with calibration strategies, and collaborate with scientists to operationalize research into production code across US/EU projects.