Senior AI Engineer, Sovereign LLMs & R&D

Home Team Science and Technology Agency (HTX); Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Home Team Science and Technology Agency (HTX) in Singapore seeks Applied AI Engineers to advance sovereign AI via LLMs and LMMs. You will align model behavior, publish models, and collaborate with a product manager and developers across guardrails and safety datasets.

The role involves building with sovereign AI tooling and distributed GPU resources. You will shape model identity and safety, publish model cards, write documentation, and contribute to joint R&D projects with global labs.

Qualifications

• Bachelor's degree in CS/DS/Engineering/Math or related field.
• Ability to explain complex technical concepts to engineers and non-technical users.
• Hands-on experience building with LLMs beyond chatbots.
• Keen to learn and experiment in a fast-paced Applied AI R&D environment.

Responsibilities

• Publish model weights and datasets on platforms like HuggingFace with release clearance.
• Co-create model cards and technical blog posts with the Product Manager for each launch.
• Write best-practice guides and run developer days for engineers and the AI community.
• Define LLM behavior, identity, tone, refusals and honesty in guidelines and training data.
• Build behavior evaluations and track them across model checkpoints.
• Run post-training experiments (SFT, RL) to improve behavior.
• Provide alignment results to inform launch decisions.

Skills

LLMs

Model alignment

Explain concepts

Documentation

Education

Bachelor's degree in CS/DS/Engineering/Math

Tools

HuggingFace

Job description

Home Team Science and Technology Agency (HTX) in Singapore seeks Applied AI Engineers to advance sovereign AI via LLMs and LMMs. You will align model behavior, publish models, and collaborate with a product manager and developers across guardrails and safety datasets.

The role involves building with sovereign AI tooling and distributed GPU resources. You will shape model identity and safety, publish model cards, write documentation, and contribute to joint R&D projects with global labs.