Software Engineer, Machine Learning Engineer (Agentic AI)
Develops autonomous agent systems that solve complex engineering problems via AI-driven exploration, optimization, and self-improvement, integrating LLM post-training and multi-agent workflows for production-grade solutions.
- Compensation
- Not disclosed
- Location
- Seoul, South Korea
- Work setup
- Not specified
- Seniority
- Not specified
- Employment
- Not specified
- Experience
- Not specified
- Category
- Ml / AI
- Posted
Skills
Must have
- Bachelor's degree in Computer Science or related field
- Experience with ML frameworks (PyTorch, JAX, TensorFlow)
- Hands-on experience implementing or experimenting with LLM or Agentic AI systems
- Ability to quickly understand research papers and translate them into practical implementations
Nice to have
- Master's or PhD in Computer Science or related field
- Experience developing agent systems (tool use, reasoning, memory, planning, multi-agent orchestration)
- Experience with LLM post-training (SFT, RL, preference optimization)
- Experience building agent evaluation systems (benchmark design, LLM-as-a-judge, automated evaluation)
- Experience with agent frameworks (LangGraph, AutoGen, Google ADK)
- Experience with distributed training, inference optimization, or large-scale experimentation
- Experience applying Agentic AI and post-training research to real-world problem-solving
Full description
About FuriosaAI
FuriosaAI builds high-performance, high-efficiency AI compute for the Inference Era. Founded in 2017 by veteran semiconductor and AI algorithm engineers, Furiosa operates globally with offices in Korea and Silicon Valley, along with a compiler-focused R&D lab in Lisbon.
Our vision is to make AI computing sustainable, enabling access to powerful AI for everyone on Earth. We solve the AI hardware energy and operational cost crisis at the architectural level, rather than through brute force, building the world's first truly AI-native compute platform to unlock the full potential of artificial intelligence for every enterprise.
About AI Transformation Team
AI Transformation Team은 AI를 기반으로 새로운 엔지니어링 문제 해결 방식과 workflow를 만들어가는 팀입니다. 다양한 AI 기술을 활용해 엔지니어링 문제를 자율적으로 탐색하고 개선할 수 있는 시스템을 개발하고 있으며, 소프트웨어, 하드웨어, 시스템을 포함한 여러 엔지니어링 영역에서 AI 기반 문제 해결 역량을 확장하는 데 집중하고 있습니다.
단순히 기존 작업을 자동화하는 것이 아니라, AI와 엔지니어가 함께 반복적인 탐색과 최적화 과정에 참여할 수 있는 새로운 접근 방식을 만들어가고 있습니다. 이를 통해 보다 다양한 가능성을 탐색하고, 반복적인 문제 해결 과정을 효율적으로 수행할 수 있도록 돕고자 합니다. 궁극적으로는 엔지니어가 반복적이고 소모적인 작업에서 벗어나 더 중요한 판단과 설계, 창의적인 문제 해결에 집중할 수 있는 방향을 지향하고 있습니다.
About the Role
이 포지션은 AI를 활용해 복잡하고 반복적인 탐색·개선 과정이 필요한 엔지니어링 문제를 해결할 수 있는 autonomous agent system을 개발하는 역할입니다.
Agentic AI를 단순한 LLM wrapper나 workflow automation으로 다루기보다, AI가 문제를 이해하고 다양한 해결 방향을 시도하며 그 결과를 바탕으로 스스로 개선해 나가는 실질적인 문제 해결 시스템으로 구현하는 데 집중합니다. 특히 고정된 형태의 agent system이 아니라, 실행 결과와 피드백을 바탕으로 지속적으로 성능과 동작 방식이 발전해 나갈 수 있는 self-improving system을 만드는 것을 중요한 방향으로 보고 있습니다.
이를 위해 agent system의 구조와 동작 방식을 설계하고, agent가 안정적으로 문제를 탐색·실행·평가할 수 있는 기반 시스템을 개발하게 됩니다. 또한 문제와 시스템의 특성에 따라 LLM post-training, agent system evaluation, 실제 엔지니어링 workflow와의 integration 등의 과제를 다루게 되며, 새로운 아이디어를 실제 동작하는 production-quality system으로 발전시키는 과정에 참여하게 됩니다.
Key Responsibilities
-
Agent planning, tool use, memory, reasoning, self-evolving mechanism 등 agent behavior 연구 및 구현
-
여러 agent가 협업하며 solution space를 탐색할 수 있는 multi-agent workflow 및 orchestration 개발
-
Agent가 생성한 결과를 자동으로 실행·검증·평가하고, 이를 바탕으로 system improvement를 반복할 수 있는 evaluation system 개발
-
LLM-as-a-judge, task-specific metric, benchmark 등을 활용한 agent evaluation 및 optimization
-
SFT, RL 등 다양한 post-training 기법 실험 및 적용
-
빠르게 변화하는 Agentic AI와 post-training 연구 흐름을 실제 엔지니어링 문제 해결에 적용 가능한 형태로 구현
Minimum Qualifications
-
컴퓨터공학 또는 유관 분야 학사 학위
-
PyTorch, JAX, TensorFlow 등 하나 이상의 ML framework를 활용한 모델 학습 또는 실험 경험