Sr Machine Learning Engineer

Blue River Technology•Santa Clara, CA

8d•$149,365 - $275,000•Hybrid

About The Position

We are seeking a Sr Machine Learning Engineer to implement deep-learning models and frameworks for scene segmentation, depth estimation, active learning, and the development of situational awareness for autonomous vehicles operating in construction and agriculture environments. You will build and evaluate machine-learning models using data from multiple sensor modalities, including vision, radar, and thermal cameras, and assess trade-offs in accuracy, robustness, and latency. This role involves researching and developing new methods to improve detection performance and increase processing speed. You will collaborate with robotics engineers to transition algorithms from desktop and server-class systems to real-time field-deployed robotic platforms, and work with systems and software engineers to design and maintain data-processing and annotation pipelines that support continuous system monitoring, evaluation, and improvement.

Requirements

Bachelor’s degree in Computer Science, Electrical Engineering, Computer Engineering, or related field plus 1 year and 6 months of related experience.
Train and optimize computer vision models for object detection, semantic segmentation, and monocular/stereo depth estimation using supervised and self-supervised learning, including loss function customization and multi-scale model training (1 yr, 6 mos).
Build and evaluate deep learning models using PyTorch and TensorFlow, implementing transfer learning, custom training loops, distributed training, gradient-based optimization, and hyperparameter tuning for large-scale image datasets (1 yr, 6 mos).
Evaluate model inference performance and runtime behavior across heterogeneous hardware platforms, including high-performance computing (HPC) GPU environments, local NVIDIA GPU development systems, and VPU-based inference on production deployment machines, to ensure real-time execution requirements are met (1 yr, 6 mos).
Design and implement end-to-end ML pipelines for data ingestion, preprocessing, model training, validation, experiment tracking, and deployment using reproducible workflows and version-controlled environments (1 yr, 6 mos).
Integrate and validate synthetic image datasets for computer vision model training, including domain alignment, data normalization, camera parameter adjustment, and evaluation of generalization performance against real-world datasets (1 yr, 6 mos).
Research and implement emerging computer vision architectures and training strategies, including transformer-based backbones, advanced loss functions, and optimization techniques, to improve model accuracy and inference efficiency (1 yr, 6 mos).
Perform dataset curation, large-scale image preprocessing, exploratory error analysis, and implement active learning strategies based on model uncertainty and diversity sampling to improve training data quality and reduce false positives (1 yr, 6 mos).

Responsibilities

Implement deep-learning models and frameworks for scene segmentation, depth estimation, active learning, and the development of situational awareness for autonomous vehicles operating in construction and agriculture environments.
Build and evaluate machine-learning models using data from multiple sensor modalities, including vision, radar, and thermal cameras, and assess trade-offs in accuracy, robustness, and latency.
Research and develop new methods to improve detection performance and increase processing speed.
Collaborate with robotics engineers to transition algorithms from desktop and server-class systems to real-time field-deployed robotic platforms.
Work with systems and software engineers to design and maintain data-processing and annotation pipelines that support continuous system monitoring, evaluation, and improvement.