Job description Who We Are Temple Allen Industries (www.templeallen.com) is at the forefront of bringing AI and Machine Learning to industrial processes for high-value assets in aerospace, marine, wind power, and transportation markets. We are currently expanding our award-winning line of Smart Automation EMMA™ systems which promise to dramatically reshape surface preparation and the robotics, machine learning, and human augmentation landscape Position: Entry Level Machine Learning Engineer We are seeking a highly skilled Machine Learning Engineer to join our dynamic team and lead projects within the Machine Learning Program. In this role, you will be responsible for completing projects associated with the training, deployment, optimization, and advancement of machine learning models that are currently running, or will be run, on the SA EMMA systems. You should be interested in the full scope of the machine learning pipeline, including data collection, annotation, simulation, training, deployment, testing, benchmarking, and model optimization. Your passion for robotics will help fuel your work in improving the EMMA robotic solution. You should be excited to show off your work, teach peers about it, and uplift your team’s skills by sharing your expertise. You should also want to be part of the design process and be excited to participate in discussions with designers, engineers, and managers to understand the system holistically and implement machine learning solutions that bring real value to the artisan and the enterprise. This role will expose you to complex and rewarding technical challenges, as well as real-world engineering and machine learning experience. You will work with a team of engineers and developers to meet the requirements of the overall EMMA system and the Machine Learning Program. Along the course of the project, mentorship and guidance will be provided to help you grow and advance your skills on both the technical and managerial fronts. You will need to be organized, systematic, and self-driven to lead projects, successfully deliver machine learning solutions that achieve system-level performance and functional specifications, and participate in discussions coordinating the Machine Learning Program’s long-term vision and objectives with other programs and major projects. In this role, you will work on major projects that create and advance the machine learning approach used to continually improve cutting-edge robotic systems that push the boundaries of technology. Requirements • Bachelor’s or Master’s degree in Machine Learning, Robotics, Computer Science, Computer Engineering, Electrical Engineering, or a related field. • Strong proficiency in modern C++ programming. • Previous experience in computer vision, machine learning, robotics, or real-time perception systems. • Previous experience training, testing, validating, and deploying machine learning models. • Experience with ROS and ROS2. • Experience with neural networks, CNNs, semantic segmentation, instance segmentation, object detection, and classification models. • Strong understanding of machine learning model architectures, including how layers, feature extractors, heads, parameters, and model size impact accuracy, latency, memory usage, and inference performance. • Experience analyzing model architecture to identify opportunities for optimization, simplification, pruning, quantization, layer reduction, or architecture tuning. • Ability to optimize models for faster inference on real-time robotic systems while maintaining acceptable accuracy, reliability, and system-level performance. • Familiarity with model optimization and deployment tools such as TensorRT, ONNX, TorchScript, OpenVINO, or similar frameworks. • Ability to implement and run machine learning models in real-time systems, edge devices, embedded systems, GPUs, or robotics platforms. • Experience benchmarking model performance using metrics such as inference time, FPS, latency, memory usage, GPU utilization, CPU utilization, and accuracy. • Experience using NVIDIA Isaac Sim or similar robotics simulation platforms for developing, testing, and validating robotic perception systems. • Familiarity with creating and configuring simulated robotic environments, including lighting, camera placement, sensor models, textures, object behaviors, aircraft geometry, and environmental conditions. • Experience generating synthetic image datasets from simulated environments to support machine learning model training, validation, and testing. • Experience creating or using RGB images, depth images, segmentation masks, annotation outputs, and other simulated sensor data for model development. • Familiarity with domain randomization techniques to improve model robustness across different lighting conditions, surface finishes, camera angles, environments, and real-world operating scenarios. • Experience comparing simulated data performance against real-world data and identifying gaps between simulation and deployment environments. • Exposure to cloud-based machine learning workflows, including training, testing, evaluating, and deploying models using platforms such as AWS, Azure, or Google Cloud. • Experience using cloud services such as AWS EC2, S3, Lambda, SageMaker, or similar tools for data storage, training pipelines, automation, and deployment. • Ability to manage large-scale datasets in cloud environments, including organizing, versioning, transferring, securing, and retrieving training data. • Familiarity with distributed training, GPU-based cloud instances, containerized machine learning workflows, and scalable model training pipelines. • Experience using Docker or similar containerization tools to support repeatable training, testing, and deployment environments. • Experience with data handling libraries and dataset preprocessing workflows. • Exposure to GPU programming, such as CUDA, is preferred. • Proficient in software development best practices, including version control systems, testing frameworks, code reviews, documentation, and maintainable software design. • Ability to create project deadlines, remain self-driven to meet those deadlines, and think critically about the long-term goals of the program. • Ability to coordinate technical work across programs and projects while aligning machine learning efforts with broader system objectives. • Ability to hold team members accountable and delegate project work efficiently. • Excellent problem-solving skills and strong attention to detail. • Eagerness to receive and implement direct feedback from the customer. • Strong written and verbal communication skills. • Ability to demonstrate strong time management skills. • Ability to work effectively in a collaborative team environment. • Ability to efficiently communicate and renegotiate requirements based on ongoing scopes of work. Responsibilities • Lead and participate in system design discussions to generate performance and functional specifications for machine learning projects. • Research different machine learning models, understand their inputs, outputs, architectures, and limitations, and determine how they can be utilized for specific EMMA system tasks. • Train, test, validate, optimize, and deploy machine learning models for use on EMMA robotic systems. • Analyze existing machine learning model architectures to understand performance bottlenecks and identify opportunities for optimization. • Modify, simplify, or remove unnecessary model layers to improve inference speed while preserving required accuracy and reliability. • Apply model optimization techniques such as pruning, quantization, layer reduction, architecture tuning, knowledge distillation, and conversion to optimized runtime formats. • Convert trained models into deployment-ready formats such as ONNX, TensorRT, TorchScript, OpenVINO, or other runtime-optimized formats. • Benchmark models before and after optimization to validate improvements in inference speed, memory usage, GPU utilization, CPU utilization, and real-time system performance. • Evaluate tradeoffs between model size, accuracy, latency, compute requirements, hardware constraints, and deployment performance. • Work with robotics and software engineers to ensure optimized models meet the timing and performance requirements of the EMMA system. • Generate datasets and annotation requirements for future models, and lead junior engineers performing annotations. • Record desired camera and sensor data from EMMA systems to use for model training, validation, and testing. • Perform data manipulation tasks including labeling, cleaning, removing outliers, organizing metadata, and splitting data into training, validation, and test datasets. • Design and implement data collection pipelines for individual client sites. • Work with the network engineer to set up databases and cloud-connected storage systems to store, organize, and sort machine learning data. • Develop and maintain simulated environments in NVIDIA Isaac Sim or similar simulation platforms to support machine learning model training, validation, and testing. • Create realistic and domain-randomized simulation scenarios that vary lighting, surface conditions, camera angles, object placement, aircraft geometry, and environmental factors. • Generate simulated RGB images, depth images, segmentation masks, and other synthetic datasets to supplement real-world data collected from EMMA systems. • Design workflows for converting simulated outputs into usable training datasets with proper labels, annotations, and metadata. • Use synthetic and real-world datasets to improve perception tasks such as semantic segmentation, object detection, classification, feature recognition, surface identification, defect detection, and sanding-region identification. • Validate machine learning models using both simulated and real-world datasets to evaluate robustness and deployment readiness. • Compare simulation-based model performance with field performance and identify where additional real-world or synthetic data is needed. • Collaborate with robotics, controls, hardware, and systems engineers to ensure simulated environments accurately represent EMMA operating conditions. • Participate in hardware selection for EMMA systems to ensure they can successfully collect data and deploy machine learning models. • Design and implement less intensive machine learning algorithms for tasks such as classification, filtering, decision support, or lightweight perception. • Execute tests validating algorithm and model performance against system-level requirements. • Summarize and trend model performance to identify missing data, failure cases, and key improvements needed for future EMMA data collection. • Train machine learning models in cloud environments using scalable GPU compute resources. • Build and maintain cloud-based data pipelines for storing, organizing, preprocessing, and retrieving machine learning datasets. • Use cloud services such as AWS S3, EC2, Lambda, SageMaker, or similar platforms to support automated model training, evaluation, deployment, and data movement workflows. • Automate recurring machine learning tasks such as dataset preprocessing, model evaluation, benchmark reporting, and synchronization between local and cloud systems. • Monitor cloud resource usage, training costs, and compute efficiency while ensuring models are trained and tested in a repeatable manner. • Generate work orders for projects and lead the team in meeting design objectives. • Assist in mentoring and guiding junior engineers and developers in their technical development. • Participate in code reviews and provide constructive feedback to improve code quality, maintainability, performance, and reliability. • Document model architecture changes, optimization decisions, benchmark results, dataset requirements, testing outcomes, and deployment recommendations. • Continuously monitor deployed model performance and identify opportunities to improve speed, reliability, accuracy, and efficiency over time. Pay Range Annual Salary: $85,000 – $100,000 The actual compensation offered will depend on a variety of job-related factors, including location, relevant education, qualifications, certifications, experience, skills, seniority, performance, and business needs. Benefits include • Paid Time off • Health Stipend Work Location Full time onsite at our Rockville, MD office location Work Authorization/Visa Sponsorship Applicants must be authorized to work for any employer in the United States, Temple Allen does not discriminate based on citizenship status or national origin. Temple Allen Industries is proud to be able to sponsor H-1B or other employment-based visas for qualified candidates for this position. Sponsorship is evaluated on a case-by-case basis and is not guaranteed. Equal Employment Opportunity Temple Allen is an Equal Opportunity Employer. Our policy is clear: there shall be no discrimination on the basis of age, disability, sex, race, religion or belief, gender, marriage/civil partnership, pregnancy/maternity, or sexual orientation. We are an inclusive organization and actively promote equality of opportunity for all with the right mix of talent, skills and potential. We welcome all applications from a wide range of candidates. Selection for roles will be based on individual merit alone.