Role Summary:
Our never-before-seen, adaptive welding products push the limit of what physical AI in general, and machine learning in particular, can do today. We are looking for an experienced and hands-on Senior Machine Learning Developer to drive perception, classification, segmentation, and advanced AI features for our NovAI intelligence suite.
As a Senior Machine Learning Developer, reporting to the Development Manager of AI/Robotics, you will lead model architecture development, decide on machine learning technology roadmaps, and serve as a critical technical leader across our AI and robotics teams.
Key Responsibilities :
- Model Development & Research
- Participate in technical interviews, onboarding, and mentoring to help build and scale the machine learning engineering team.
- Build, train, and evaluate computer vision/ML models (detection, segmentation, tracking, pose estimation, image fusion) using PyTorch and TensorFlow.
- Fine-tune neural network architectures (CNN, RNN, Transformer) and classical ML methods for real-time welding process sensing.
- Define short and long-term machine learning technology roadmaps, lead fast prototyping, and select AI tools and framework architectures.
- Implement vision-language-action (VLA) models and foundation architectures for spatial reasoning and multi-modal perception.
- Develop and integrate world models for predictive perception, spatial reasoning, and dynamic environmental simulation in adaptive welding scenarios.
- Develop closed-loop reinforcement learning (RL), imitation learning, and continuous learning policies for complex control tasks, adaptive manipulation, and robotic feedback loops.
- Architect simulation-to-real (Sim2Real) pipelines and synthetic data workflows (Isaac Gym, MuJoCo) to accelerate domain randomization
- Physical AI & Robotics Integration
- Fuse vision, arc, and thermal sensor data to enable real-time robotic decision-making.
- Integrate ML outputs with motion control, PLCs, and robot controllers within target latency, power, and compute budgets.
- Support toolpath and trajectory optimization for stress reduction, energy efficiency, and faster cycle times.
- Deployment, Testing & MLOps
- Optimize models for real-time edge implementation and maximize hardware efficiency using quantization, pruning, TensorRT, and ONNX Runtime on embedded platforms.
- Validate and troubleshoot end-to-end integrations across software and hardware stack components.
- Maintain state machine documentation, architecture diagrams, technical specifications, and user guides.
- Provide onsite and remote commissioning support during customer installations.
- Implement robust MLOps practices, including automated model versioning, validation workflows, and monitoring (W&B/MLflow).
- Cross-Functional Collaboration
- Partner with controls and welding engineering teams to integrate robotic systems with external automation hardware.
- Get to the root of one-time and repeating integration or customer Service issues and implement long-term, sustainable solutions.
- Contribute to scientific publications and patent applications; evaluate emerging ML tooling and frameworks.
- Travel as required to customer deployment sites or industry conferences.
Qualifications and Experience :
- Bachelor's degree in CS, Electrical Engineering, Software Engineering, or related field (Master's preferred).
- 5+ years of practical experience with PyTorch, TensorFlow, and OpenCV.
- 3+ years of experience with deep learning frameworks (TensorFlow, Keras, PyTorch).
- Strong foundation in supervised/unsupervised learning, deep neural networks (CNNs/RNNs), and transfer learning.
- Solid computer vision skills: image processing, object detection, segmentation, and feature extraction.
- Proficiency in C++ and Python applied to robotic systems.
- Strong technical documentation, communication, and Git-based workflow skills.
- Strategic, critical thinking, and analytical skills for structured problem-solving rather than band-aid fixes
Preferred Additional Qualifications:
- Experience integrating an AI pipeline with a third party PC/PLC
- Hands-on exposure to edge inference optimization or Physical AI systems.
- Familiarity with industrial sensing or automated welding processes
Commitment to Novarc Values:
Demonstrate and embed Novarc’s core values of Hungry, Humble, and Smart in daily work, decision-making, collaboration, and interactions with colleagues, customers, and partners.
- Hungry: Be Passionate, Never Give Up, Customer Focused, Continuous Learning, Do What You Say
- Humble: Servant Leader, No arrogance, Less Ego, Listen > Talk
- Smart: High EQ, Challenge the Status Quo, Thought Leaders
