Analog & Mixed-Signal Design
$172,000 - $236,500 USD yearly
Originally posted November 16, 2026 by the employer.
About the role
This role involves collaborating with engineering teams to collect datasets and train machine learning models that execute directly on sensing devices. The focus is on developing a sensing platform tailored for data centers.
What you'll do
- Define model training strategy, including dataset selection, splitting, validation methods, and data volume requirements from lab and field environments.
- Build physical and ML models to generate operational insights and recommendations.
- Own model evolution by managing versioning, tracking drift, retraining across site rollouts, and evaluating deployment readiness for edge units.
- Align with peer modeling engineers to standardize models, feature sets, and evaluation methods across teams.
- Establish data hygiene and quality standards across labeling, calibration records, sensor drift monitoring, and schema conformance.
- Communicate model performance, including limitations and edge cases, to product, hardware, and firmware leads.
What you'll need
- MS in engineering, physics, applied math, or computer science, and 8+ years building models on real sensor data.
- Strong proficiency in Python and the standard ML stack (scikit-learn, PyTorch, or similar).
- Experience handling time series, anomaly detection, and small, noisy datasets.
- Experience combining first-principles or physics-based models with data-driven models.
Skills: ML model, sensor data, Python, PyTorch, time series
More semiconductor jobs: Analog Devices jobs · MEMS jobs
Wilmington, MA, United States
On-site
Full-time
$172,000 - $236,500 USD yearly
2 days ago