About GovTech, Singapore
The Government Technology Agency (GovTech) is the leading agency behind Singapore’s Smart Nation initiatives, focusing on public sector digital transformation. As the Centre of Excellence for Infocomm Technology and Smart Systems (ICT & SS), GovTech enhances the Singapore Government’s capabilities in Data Science & AI, Application Development, Smart City Technology, Digital Infrastructure, and Cybersecurity.
Role Overview:
As an AI Engineer in the Responsible AI team, you will conduct safety testing of large language models (LLMs) and develop safety guardrails. Your work will involve collaborating with AI product teams to ensure responsible AI practices.
Key Responsibilities
- Execute safety testing on LLM systems and develop corresponding safety guardrails.
- Experiment with the latest open-source tools and create your own solutions, applying them to real-world LLM products launched by the government.
- Fine-tune new models for safety alignment to support internal safety requirements.
- Collaborate closely with AI product teams and public officers to deepen their understanding of responsible AI and integrate safety testing and guardrails into their products.
- Engage with various technical partners from the private sector, academia, and international organizations to foster collaborative innovation.
- Share insights and findings with the community, showcasing GovTech’s leadership in LLM safety testing and guardrails.
Qualifications
- Technical Skills: Strong backend programming skills in Python, particularly in API development and ML performance optimization.
- Cloud Experience: Demonstrated experience with cloud deployments, especially on AWS, to ensure scalable and efficient AI solutions.
- Understanding of LLMs: A solid technical grasp of how LLMs operate, effective prompt engineering techniques, and app development using these models.
- Interest in AI Safety: A genuine interest in AI safety and responsible AI, with a proactive approach to learning and experimentation.
- Creative Problem-Solving: Ability to think creatively and independently to solve challenges and drive innovation.
- Public Good Orientation: A strong desire to develop technology that serves the public good.
Preferred Experience
Candidates with informal or formal experience in the following areas will be highly regarded:
- Conducted LLM safety testing, including jailbreaking or prompt injections.
- Developed guardrails for LLMs or use equivalent tools (e.g., NeMo, LLM-Guard).
- Built and deployed high-performance LLM-based applications.
Employee Benefits
- A comprehensive total rewards approach, offering a holistic and competitive suite of benefits, including generous leave policies to support work-life balance.
- Wellness programs to promote employee health and well-being.
- Flexible work arrangements that allow employees to manage their own time and work environment effectively.
To apply for the job, click here: Careers (myworkdayjobs.com)