AI Research Engineer Computer Vision & Analytics

Key Responsibilities

• Model Training and Development: Lead the complete training pipeline for computer vision based analytical models, including data preparation, augmentation, and implementing custom training and fine-tuning routines to meet specific project goals.
• Custom Model Adaptation: Develop and customize model architectures to accurately identify unique and specific needs required by our applications.
• Accuracy and Performance Enhancement: Continuously iterate on and experiment with models to systematically improve key metrics such as precision, recall, and mAP (mean Average Precision).
• Performance & Cost Optimization: Continuously analyze and optimize cloud resource consumption (e.g., CPU, GPU, memory, and storage) to ensure cost-effective, high-throughput operations and meet budget targets.
• Accelerator Implementation: Identify, benchmark, and leverage the most appropriate cloud hardware and software accelerators (e.g., specific GPU/TPU instances, inference compilers like NVIDIA TensorRT, AWS Inferentia) to enhance processing speed and reduce latency.
• Deployment and Integration: Work closely with software and hardware engineering teams to ensure the seamless integration and deployment of optimized models into our final products.
• Performance Benchmarking: Establish and execute robust testing protocols to measure and report on key performance indicators (KPIs), including inference latency, throughput, accuracy, and cost-per-inference/cost-per-query.
• Algorithm Development: Develop analytical algorithms for a variety of use cases, with an emphasis on deep mathematical and theoretical principles.

Required Qualifications

• Bachelor's or Master's degree in Computer Science, Electrical Engineering, Artificial Intelligence, or a related technical field.
• Strong programming skills in Python.
• Proven experience with modern deep learning frameworks such as PyTorch (preferred) or TensorFlow.
• Solid foundation in computer vision principles and deep learning concepts.
• Hands-on experience with training and fine-tuning machine learning models.
• Prior contributions to research papers or projects.

Preferable Skills and Experience - Machine Learning

• Direct, hands-on experience with the NVIDIA Jetson family of devices (e.g., Orin, Xavier, Nano) and its development environment.
• Demonstrated expertise in model optimization and acceleration using NVIDIA TensorRT.
• In-depth knowledge of the YOLO architecture (e.g., YOLOv5, YOLOv8, YOLO-NAS) and experience in custom training it.
• Proven experience with model pruning tools and techniques.
• Familiarity with MLOps principles and tools (e.g., Docker, MLflow) for managing the machine learning lifecycle.
• Experience with video pipeline tools like GStreamer or NVIDIA DeepStream.
• A portfolio of relevant projects, a GitHub profile showcasing your work, or contributions to open-source projects.

Preferable Skills and Experience - Mathematical

• Strong foundation in calculus, including:
• Differentiation (e.g., gradients, Jacobians, Hessians) for optimization in machine learning.
• Integration (e.g., continuous probability distributions, loss functions).
• Linear algebra (e.g., matrix operations, eigenvalues, singular value decomposition).
• Probability and statistics (e.g., Bayesian inference, statistical modeling).
• Optimization techniques (e.g., gradient descent, convex optimization).

Soft Skills

• Strong problem-solving abilities and attention to detail.
• Excellent communication skills for documentation and collaboration.
• Ability to work effectively in a team-oriented environment.

What We Offer

• Exceptional Peer Environment: The opportunity to work directly with and learn from experts in the field of AI.
• Impactful Work: Contribute to fundamental research and have a lasting impact on the industry.
• State-of-the-Art Resources: Access to datasets and extensive computational resources to bring your most ambitious ideas to life.
• Culture of Growth: A dynamic and intellectually stimulating environment that encourages continuous learning and professional development.