Meta is seeking a Software Engineer III on a one-year contract assignment in New York, United States, for work focused on artificial intelligence, machine learning, and large language model (LLM) evaluations. The position is an on-site opportunity within a research laboratory environment and is being offered through Tundra Technical Solutions, a staffing and consulting services provider. The successful candidate will join Meta’s Pretraining Research Team within the Fundamental AI Research (FAIR) organisation. The team works on evaluating large language models across different stages of development, including pretraining, midtraining, and post-training. The role is particularly suited to professionals with experience in LLM-related research and evaluation, especially those who have worked on pretraining evaluations. Software Engineer III Role and Research Focus The Software Engineer III will contribute to the development and execution of LLM evaluation processes designed to measure model performance and capabilities. The position requires candidates who can work with large volumes of technical evaluation data and translate results into useful insights for AI research teams. Key responsibilities include: Running LLM evaluations during pretraining, midtraining, and post-training. Organising evaluation results across multiple models, benchmarks, and metrics. Porting evaluations between different evaluation frameworks and codebases. Ensuring the correctness and reproducibility of evaluation results. Developing, testing, and improving prompts used in evaluation benchmarks. Analysing evaluation findings across different models and benchmarks. Creating clear summaries and visualisations to communicate research findings. Supporting research teams in making decisions based on evaluation results. The successful candidate will therefore combine software engineering expertise with practical knowledge of machine learning and LLM evaluation methodologies. Minimum Qualifications and Technical Skills Applicants are expected to have between three and five years of experience working on LLM-related topics, including pretraining and pretraining evaluation. The minimum requirements include: A bachelor’s degree in Computer Science or a related technical field. Three to five years of relevant experience involving LLMs and related technologies. Experience with Python. Experience with PyTorch. Experience building systems using machine learning and/or deep learning methods. Knowledge of LLM evaluation, particularly pretraining evaluation. Candidates with a strong combination of software engineering and machine learning experience may be well positioned for the role. Preferred Knowledge of LLM Evaluation Meta identifies several areas of knowledge that would be advantageous for applicants. One is an understanding of perplexity, including what the metric measures, how it behaves across different model sizes, its relationship with downstream task performance, and its limitations. Familiarity with scaling ladders is also desirable. This includes experience with training models of different sizes and understanding how evaluation metrics can change as model scale increases. These preferred qualifications indicate that the role requires more than general software development experience. Candidates are expected to understand the technical principles behind evaluating and comparing modern AI models. Contract Details and Compensation The Software Engineer III position is a one-year contract assignment based in New York and requires on-site work. The position is not a full-time Meta employment opportunity. Successful candidates will be engaged as contractors through Tundra Technical Solutions for the specified duration of the assignment. The advertised pay rate ranges from $90 to $98.72 per hour in United States dollars. Tundra Technical Solutions states that its benefits offering may include optional medical, dental and vision benefits, retirement benefits, up to 15 days of paid time off per year, and a New Child Benefit. Candidates seeking clarification about contract terms or the engagement arrangement are advised to contact Tundra Technical Solutions directly. Work Environment and AI Research Opportunities The successful candidate will work within a laboratory environment dedicated to AI research and will collaborate with researchers working on large-scale model evaluations. The position offers exposure to several areas of modern AI development, including pretraining evaluation, prompt engineering, benchmark analysis, model scaling, machine learning systems, and research data visualisation. The role may therefore be particularly relevant to software engineers and machine learning professionals seeking experience at the intersection of software engineering and advanced AI research. Application Considerations Interested candidates should carefully review the stated technical requirements before applying. Experience with Python and PyTorch, combined with practical experience working with LLMs and machine learning systems, is central to the position. Applicants should also ensure that their CV demonstrates relevant experience in LLM evaluation, pretraining, machine learning or deep learning systems, software development, and research-oriented technical projects where applicable. The recruitment process is subject to applicable employment and fair-chance laws. Tundra Technical Solutions states that qualified applicants with arrest or conviction records will be considered in accordance with applicable legislation. The Software Engineer III opportunity provides an opportunity for an experienced technical professional to contribute to LLM evaluation work within Meta’s AI research environment while working on a contract basis through Tundra Technical Solutions. VISIT OFFICIAL WEBSITE TO APPLY For more opportunities such as these please follow us on Facebook, Instagram, Twitter, LinkedIn and WPChannel Disclaimer: Global South Opportunities (GSO) shares jobs, internships, fellowships, grants, scholarships, and other opportunities for informational purposes only. GSO is not affiliated with Magamba Network or the Fired Up initiative and does not participate in the recruitment or selection process. Applicants should verify the vacancy details, eligibility requirements, application process, and deadline through the official organisation before applying.