PHYSICS-INFORMED MACHINE LEARNING ENGINEER

Svitla Systems, Inc.; Argentina; Presencial; ARS 2.500.000 - 4.500.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Private medical insurance

Flexible workspace (remote or office)

Bonuses for referrals

Bonuses for article writing and talks

15 vacation days, 10 holidays, 10 sick

Descripción de la vacante

Svitla Systems, Inc. is seeking a Physics-Informed Machine Learning Engineer for a full-time role in Argentina.

You will build physics-informed models to predict degradation and health, using time-series data from sensors and telemetry, while combining physics-based losses with data-driven optimization. You will design a fusion layer that blends physics stress features with dynamical signals, and you’ll calibrate stress-proxy parameters and maintain the end-to-end data pipeline in Python.

Formación

• Experience in building and training physics-informed models — a physics-based term in the loss function of a real project (PINN, physics-regularized NN, or equivalent).
• Strong understanding of time-series/sequence modeling (LSTM, temporal CNN, transformers, or state-space models) on sensor or telemetry data.
• Understanding of parameter calibration / inverse problem: fitting mechanistic model parameters to noisy observational data (Bayesian calibration, MLE, or optimization-based).
• Expert knowledge of Python scientific stack (Pandas, NumPy, scikit-learn, PyTorch or JAX) and be comfortable owning a data pipeline end to end, including data-quality investigation.
• Expertise in reading and reasoning about physics/reliability equations governing degradation; you don’t need to derive them, but they can’t be a black box.

Responsabilidades

• Build the temporal model: design and train a physics-informed sequence model for degradation and health prediction, incorporating a physics-based loss term alongside the data-driven loss.
• Design the fusion layer: define how physics-based stress features, dynamical/mathematical features, and other signals combine into model inputs and a defensible health score.
• Calibrate the physics-informed components: the stress-proxy parameters are currently engineering priors; you’ll help design and execute calibration strategies against available outcome labels.
• Harden the feature pipeline: the pipeline is Python/Pandas over time-aligned multi-sensor telemetry; extend and maintain it as modeling needs dictate.
• Write clear analysis docs and defend modeling choices to technical stakeholders and clients.

Conocimientos

Physics-informed ML

Time-series modelling

Bayesian calibration

Python data stack

Model interpretability

Educación

Master's degree in related field

Herramientas

PyTorch

JAX

Pandas

NumPy

scikit-learn

Descripción del empleo

Svitla Systems Inc. is looking for a Physics-Informed Machine Learning Engineer for a full-time position (40 hours per week) in Argentina.

Requirements

• Experience in building and training physics-informed models — a physics-based term in the loss function of a real project (PINN, physics-regularized NN, or equivalent).
• Strong understanding of time-series/sequence modeling (LSTM, temporal CNN, transformers, or state-space models) on sensor or telemetry data.
• Understanding of parameter calibration / inverse problem: fitting mechanistic model parameters to noisy observational data (Bayesian calibration, MLE, or optimization-based).
• Expert knowledge of Python scientific stack (Pandas, NumPy, scikit-learn, PyTorch or JAX) and be comfortable owning a data pipeline end to end, including data-quality investigation.
• Expertise in reading and reasoning about physics/reliability equations governing degradation; you don't need to derive them, but they can't be a black box.

Nice to have

• Experience in reliability engineering/PHM (prognostics and health management) background: RUL estimation, degradation modeling, accelerated-life testing.
• Exposure to semiconductor or hardware degradation physics at a "read the literature critically" level.
• Familiarity with nonlinear dynamics/recurrence or dynamical-systems features (e.g., RQA or comparable techniques).
• Familiarity with hardware/datacenter telemetry or fleet analytics.
• Experience working in small teams alongside domain scientists/mathematicians; comfortable turning research feedback into production code.

Responsibilities

• Build the temporal model: design and train a physics-informed sequence model (e.g., LSTM or similar temporal architecture) for degradation and health prediction, incorporating a physics-based loss term alongside the data-driven loss.
• Design the fusion layer: define how physics-based stress features, dynamical/mathematical features, and other signals combine into model inputs and a defensible health score, replacing today's simple hand-set weighting.
• Calibrate the physics-informed components: the stress-proxy parameters are currently engineering priors. You'll help design and execute calibration strategies against whatever outcome labels are available.
• Harden the feature pipeline: the pipeline is Python/Pandas over time-aligned multi-sensor telemetry; you'll extend and maintain it (feature audits, label engineering, data-quality gates) as modeling needs dictate.
• Write clear analysis docs and defend modeling choices to technical stakeholders and clients.

WE OFFER

• US and EU projects based on advanced technologies.
• Competitive compensation based on skills and experience.
• Comprehensive private medical insurance.
• Regular performance appraisals to support your growth.
• Flexibility in workspace, either remote, our welcoming office or local coworking.
• Bonuses for recommendations of new employees.
• Bonuses for article writing, public talks, other activities.
• 15 vacation days, 10 national holidays, 10 sick leaves.
• Personalized learning program tailored to your interests and skill development.
• Free tech webinars and meetups organized by Svitla.
• Fun corporate online/offline celebrations and activities.
• Well-established remote culture.
• Awesome team, friendly and supportive community!