AI Safety Engineer

<p><strong>Sustainable Talent </strong>is partnering with <strong>Nvidia</strong> a global leader who's been transforming computer graphics, PC gaming, and accelerated computing for over 25 years. We are looking for a <strong>Machine Learning Engineer- AI Safety & Security</strong> to support our client's team based out of in <strong>Santa Clara, CA with remote/ hybrid work options.</strong> </p> <p> </p> <p>This is a full-time (W-2) <strong>contract</strong> role. We offer competitive pay <strong>$90/hr - $130/hr </strong>based on factors like experience, education, location, etc. and provide full benefits, PTO, and amazing company culture!</p> <p>As a Machine Learning Engineer, you'll work alongside NVIDIA’s research and engineering teams, focused on AI Safety for LLMs, including multi-lingual, multi-modal, and reasoning models.  We value expertise in data science paired with a robust data engineering foundation.   This role is directed at assessing, and improving the safety and inclusivity of our LLM models in a scalable fashion.  <strong>We seek someone proficient in programming and scripting for comprehensive data manipulation, analysis, and model fine-tuning.</strong>  We believe in proactive problem-solving, minimal supervision, and being exceptional teammates who collaborate, think, and learn as one unit. Let's make a difference together!</p> <p><span style="text-decoration: underline;"><strong>What you’ll be doing:</strong></span></p> <ul> <li>Develop datasets and moderator models for evaluating LLM models and end-to-end systems for Content Safety, ML Fairness. These LLM models can be txt-to-txt or multimodal-to-txt.</li> <li>Develop datasets for training LLM models with SFT and RL techniques, for Content Safety, ML Fairness, Security and more.</li> <li>Research and implement cutting-edge techniques for bias detection and mitigation in LLMs and systems.</li> <li>Define and track key metrics for responsible LLM behavior and usage.</li> <li>Follow the best practices of automation, monitoring, scale, safety.</li> <li>Contribute to our repositories and develop safety tools to help ML teams be more effective.</li> <li>Data pre-processing and analysis: Collaborate with data scientists and data engineers to collect, clean, pre-process, and transform large and wide datasets.</li> <li>Conduct exploratory data analysis (EDA) to uncover insights and identify patterns that boost the model performance.</li> <li>Collaborate with multidisciplinary teams: Collaborate with product engineers, data scientists, and analysts to understand business requirements and translate them into