Intermediate Machine Learning Developer

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.

Prototype and test new features against the applied research roadmap defined by senior engineering.

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.

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.

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).

Partner with controls and welding engineering teams to integrate robotic systems with external automation hardware.

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).

3+ years of practical experience with PyTorch, TensorFlow, and OpenCV.

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 communication and documentation skills

Strong organizational and time management 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