About the job
Sustainable Talent is collaborating with Nvidia, a pioneering organization that has been reshaping computer graphics, PC gaming, and advanced computing for more than 25 years. We are in search of a Machine Learning Engineer specializing in LLM Safety & Security to join our client's team located in Santa Clara, CA, with flexible remote or hybrid work arrangements.
This is a full-time (W-2) contract position. We offer competitive compensation ranging from $90/hr to $130/hr based on experience, education, location, and other relevant factors, along with comprehensive benefits, paid time off, and an exceptional company culture!
In your role as a Machine Learning Engineer, you will collaborate with NVIDIA’s research and engineering teams, concentrating on AI Safety for LLMs, including multilingual, multimodal, and reasoning models. We value a strong foundation in data science complemented by expertise in data engineering. This position is aimed at evaluating and enhancing the safety and inclusivity of our LLM models at scale. We are looking for someone skilled in programming and scripting for thorough data manipulation, analysis, and model optimization. Our approach emphasizes proactive problem-solving, minimal supervision, and being outstanding team players who collaborate effectively, think critically, and learn collectively. Let’s create impactful solutions together!
Key Responsibilities:
- Design datasets and moderator models for assessing LLM models and complete systems regarding Content Safety and ML Fairness, including text-to-text and multimodal-to-text models.
- Create datasets for training LLM models utilizing SFT and RL techniques focused on Content Safety, ML Fairness, and Security.
- Investigate and apply state-of-the-art methods for bias detection and mitigation within LLMs and associated systems.
- Establish and monitor key metrics for responsible LLM behavior and usage.
- Adhere to best practices in automation, monitoring, scaling, and safety.
- Contribute to our code repositories and develop safety tools to enhance the effectiveness of ML teams.
- Engage in data preprocessing and analysis: Collaborate with data scientists and engineers to gather, clean, preprocess, and transform extensive datasets.
- Conduct exploratory data analysis (EDA) to reveal insights and patterns that enhance model performance.
- Work in partnership with multidisciplinary teams: Collaborate with product engineers, data scientists, and analysts to comprehend business needs and translate them into actionable strategies.

