AI/ML Engineer

Latent AI•Princeton, NJ

4d•Onsite

About The Position

Latent AI is seeking a hands-on machine learning engineer to build and ship production ML systems, including ML infrastructure developed from scratch. The role requires sound judgment in selecting efficient, reliable models and understanding the relevant data modalities, including audio, vision, and speech. The engineer will improve physical AI inference pipelines and collaborate with the team and customers on solutions that can be deployed in the field. As the company’s ML capability grows, the engineer will also mentor teammates and raise the company’s ML standards. The position is based in Princeton, New Jersey, with a preference for local, in-office work. U.S. citizenship is required due to the nature of the projects.

Requirements

3–5+ years of machine learning engineering experience on a small team, with a strong track record of shipping computer vision or ML systems to production.
Hands-on experience building perception and autonomous-systems solutions, such as computer vision.
Experience building ML infrastructure from scratch, including training pipelines, serving infrastructure, evaluation harnesses, and monitoring.
Startup or small/mid-size company experience, with ownership of meaningful scope and pragmatic build-versus-buy-versus-defer decisions.
Strong understanding of feature engineering, leak-aware train/test splits, metric selection for imbalanced data (including precision, recall, F1, and AUC rather than relying on accuracy alone), cross-validation, and principled hyperparameter tuning.
Strong Python skills and hands-on experience with AI/ML frameworks such as PyTorch and scikit-learn.
Solid MLOps foundation, including ML CI/CD, model versioning, experiment tracking, drift detection, and production monitoring.
Experience designing inference systems with explicit latency and throughput targets on physical edge hardware (CPU/GPU), and making informed tradeoffs among model complexity, cost, and performance.
U.S. citizenship, required due to the nature of the projects.

Nice To Haves

Experience with AWS.
Experience training, evaluating, and serving models.

Responsibilities

Prepare real-world data for modeling and select an appropriate model approach, from classical classifiers to fine-tuned LLMs or VLMs, based on tradeoffs such as cost and latency.
Improve the physical AI inference pipeline by deliberately balancing performance, latency, and accuracy.
Partner with the team to engage customers and build solutions deployable in the field.
Mentor teammates and raise the company’s ML standards as its ML capability matures.