Google DeepMind is inviting applications for the position of Research Engineer, Polaris in Mountain View, California, United States. The role offers an opportunity for software developers and technically skilled professionals to contribute to the evaluation and improvement of Google’s Gemini artificial intelligence models. The successful candidate will join the Polaris team, which develops challenging software engineering tasks to assess how effectively Gemini can solve complex, real-world programming problems. The work involves designing realistic evaluation tasks, developing scoring rubrics and supporting projects that improve AI training data and evaluation systems. The advertised US salary range is US$147,000 to US$210,000, with a target annual bonus of 15%, potential equity awards and additional employee benefits. Individual compensation depends on job-related skills, experience, education and training. The position is based in Mountain View, California. Applicants should review the official Google Careers listing for current application instructions and any additional requirements. About Google DeepMind and the Polaris Team Google DeepMind is an artificial intelligence research laboratory focused on advancing AI capabilities and developing technologies intended to benefit society. Its work spans areas such as machine learning, scientific discovery, language processing and AI-enabled applications. The Polaris team focuses on evaluating Gemini’s ability to complete long-horizon software engineering tasks. These tasks test whether AI models can handle complex programming assignments that require sustained reasoning, implementation and problem-solving rather than simply answering short coding questions. Each task includes a prompt describing the required work and a grader or rubric that assigns a numerical score to the model’s attempted solution. The team places particular emphasis on manually creating a smaller number of realistic, technically demanding tasks instead of relying primarily on large volumes of synthetic tasks. According to the job description, developing a single well-designed task can take approximately one week, from initial idea to final approval. Research engineers therefore need strong programming skills, technical judgement and the ability to design meaningful tests of AI performance. Position Details Job title: Research Engineer, Polaris, DeepMind Employer: Google DeepMind Location: Mountain View, California, United States Salary range: US$147,000–US$210,000 annually Additional compensation: 15% target bonus, potential equity and employee benefits Work area: Artificial intelligence, software engineering and model evaluation Application deadline: Not specified in the supplied listing The advertised compensation is for the United States. Applicants should consult the official job page for details about employment arrangements, the role’s current status and any additional conditions. Key Responsibilities The Research Engineer will contribute to the development of evaluation tasks and projects designed to measure and improve Gemini’s software engineering capabilities. Responsibilities include: 1. Designing complex software engineering tasks The engineer will create tasks that assess Gemini’s ability to complete extended programming assignments. These tasks should reflect realistic technical challenges and require meaningful problem-solving. 2. Developing evaluation criteria Each task needs an appropriate grader or rubric that can assess the quality of the model’s output. The engineer will help design evaluation methods that produce useful and consistent measurements of model performance. 3. Supporting AI training and evaluation projects The role involves launching new projects within DeepMind that aim to improve Gemini’s training data and evaluation systems. These initiatives help identify model weaknesses and provide information that can guide future development. 4. Applying software engineering expertise The successful candidate will use programming experience and technical judgement to develop tasks that are sufficiently challenging, realistic and measurable. 5. Learning new tools and technical approaches The position favours candidates who can adapt quickly to unfamiliar programming paradigms, development tools and technical stacks. Experience with AI-assisted coding tools and developer agents is particularly relevant. Minimum Qualifications Applicants must demonstrate the following: Experience in software development or programming. Experience using AI-assisted coding tools, developer agents or large language model-based workflows for software development. Practical experience using coding agents to perform work. The emphasis on hands-on experience with coding agents is important. Applicants should be able to explain how they have used these tools in actual development workflows rather than relying only on general familiarity with AI. Preferred Qualifications The following experience may strengthen an application: Participation in competitive programming or mathematics competitions, such as the International Mathematical Olympiad, USAMO, Putnam or International Olympiad in Informatics. Participation in hackathons. Experience designing evaluation rubrics, benchmark datasets or assessment frameworks for machine learning models. Ability to learn new programming languages, tools, technical stacks and development paradigms quickly. Candidates do not necessarily need experience in every preferred area. They should focus on demonstrating strong programming ability, effective problem-solving and relevant experience with AI-assisted software development. Salary, Bonus and Benefits Google advertises an annual salary range of US$147,000 to US$210,000 for this position. The compensation package may include: Annual base salary: US$147,000–US$210,000. Target annual bonus: 15% of the applicable compensation basis, subject to the company’s bonus arrangements. Equity: Potential stock-based compensation, subject to approval and the applicable grant agreement. Employee benefits: Additional benefits described on Google’s official careers benefits page. Equity awards are discretionary and are not guaranteed. Actual compensation may vary depending on qualifications, experience, education, training and other job-related factors. For further information, applicants can review Google’s employee benefits. How to Strengthen an Application Candidates should tailor their CVs to demonstrate their ability to build software, solve difficult programming problems and use AI-assisted development tools effectively. Where possible, applicants should provide concrete examples of projects involving coding agents, LLM-based development workflows, automated testing, benchmark design or evaluation of AI-generated code. Experience in competitive programming, mathematics competitions or hackathons may help demonstrate algorithmic thinking and the ability to solve unfamiliar technical problems. Applicants with experience designing rubrics or test datasets should explain how their evaluation methods measured performance and identified weaknesses. A strong application should also show an ability to work independently, adapt to new tools and translate complex technical requirements into well-defined tasks with measurable outcomes. How to Apply Interested candidates should apply through the official Google Careers website. Review the job description carefully and submit the requested application materials through the online recruitment system. Applicants should verify that the vacancy is still open and review the full requirements before submitting their applications. Related Opportunity: Evals Engineer, Polaris Google also lists an Evals Engineer, Polaris, DeepMind (Fixed-Term Contract) position in Mountain View. Candidates interested in AI model evaluation may wish to review this separate vacancy as well. VIEW THE RELATED EVALS ENGINEER POSITION Why This Opportunity Is Valuable This position offers experienced programmers an opportunity to work on the evaluation of advanced AI systems while contributing to the development of more capable software engineering models. The role combines practical programming, task design, benchmarking and AI research, making it relevant to candidates interested in the intersection of software engineering and artificial intelligence. The competitive advertised salary, target bonus, potential equity and employee benefits make it a notable opportunity for qualified technical professionals seeking to work with Google DeepMind in the United States. 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) does not take ownership or credit for the opportunities shared on this platform. All opportunities are sourced from official announcements and external sources for informational purposes only. Applicants are encouraged to verify all details, deadlines, and eligibility requirements from the official source before applying. Opportunities may be updated or changed by the providing organisations.