Senior Machine Learning Engineer

This is a fully remote position, open to applicants in Canada.

📋 Description

• Develop and refine both large and small language models specifically for game-agent applications.

• Create and sustain data preparation and training workflows.

• Curate training examples and assess data quality.

• Ensure a clear distinction between training and evaluation datasets.

• Design reproducible experiments that include hypotheses, baselines, and evaluation metrics.

• Evaluate candidate models based on capability, reliability, latency, memory usage, and inference expenses.

• Investigate model failures in collaboration with the Agents and Game Technology teams.

• Utilize representative scenarios and diagnostic insights to enhance training data, model performance, or agent interfaces.

• Collaborate with Platform engineers to prototype and benchmark models using serving frameworks like vLLM and NVIDIA Triton.

• Transform relevant research findings into practical implementations.

• Document methodologies, results, and limitations.

• Communicate insights through code reviews and provide technical guidance.

• Lead model development initiatives, experimental evidence, and reusable training workflows.

• Work together on production integration, validation, and troubleshooting.

⛳️ Requirements

• An AI-first mindset in engineering, with practical experience utilizing AI tools for development, testing, debugging, or analysis.

• Over 5 years of experience in machine learning engineering or a closely related field, including significant hands-on experience in fine-tuning and deploying language models.

• Proficient programming skills in Python along with experience in PyTorch, JAX, TensorFlow, or similar machine learning frameworks.

• Practical knowledge of language model training and post-training techniques, including supervised fine-tuning, preference optimization, or reinforcement learning.

• Experience in creating reproducible experiments, preparing training data, evaluating models, and identifying regressions.

• Solid understanding of model-performance trade-offs, including latency, throughput, GPU memory, inference costs, and overall model quality.

• Familiarity with model-serving frameworks such as vLLM, NVIDIA Triton, or equivalent tools.

• Competence in sound software engineering practices, including testing, version control, code review, and maintaining reusable tools.

• Ability to independently conduct technical investigations and articulate findings and trade-offs to researchers, engineers, and game developers.

• A Master’s or PhD in Computer Science, Machine Learning, Artificial Intelligence, or a related discipline is preferred; equivalent practical experience is also valued.

• Experience with multimodal models, model distillation, quantization, or on-device inference would be advantageous.

• A genuine interest in gaming and the impact of agent behavior on player experience.

• Comfort in a dynamic startup environment.

🏝️ Benefits

• Engage in innovative projects that push the boundaries of AI in gaming.

• Collaborate with a talented team in a fast-paced environment.

• Opportunities for professional development and growth.

• Competitive salary and benefits package.