(Seoul) Applied Research Scientist · AI Innovation - 전문연 가능

"Conquering cancer through AI"Lunit, a portmanteau of ‘Learning unit,' is a medical AI software company devoted to providing AI-powered total cancer care.Our AI solutions help discover cancer and predict cancer treatment outcomes, achieving timely and individually-tailored cancer treatment.🗨️ About The TeamWho will I spend 8+ hours/day with?You will join the AI Innovation team behind the Chain-of-Evidence (CoE) project, also known as 특화 파운데이션 모델 project. The team brings together AI research scientists and engineers across RAG & Product, Model & Evaluation, and Data & Knowledge, building real products while tackling meaningful applied research along the wayMembers have diverse backgrounds and interests — some are passionate about LLM training and evaluation, others about retrieval and knowledge systems, others about productization and clinical deployment — united by a shared commitment to improving patient careThis team is a fast paced "full stack" team, where ideation to deployable PoC happens in rapid iteration cycles. The team gets to work on strategically important initiatives aligned to enabling new business opportunities for Lunit🗨️ About The PositionWhat will make me proud to work here?You will have a direct impact on bringing trustworthy, evidence-grounded AI into real clinical workflows — clinical intelligence and agentic applications — deployed and evaluated together with partner hospitalsYour work will push forward the performance of Lunit's medical foundation models, RAG systems, and agentic pipelines that power Lunit's next generation clinical intelligenceWe work with large-scale clinical knowledge sources (medical literature, clinical guidelines, insurance and EMR data) and the cloud/GPU compute to leverage themCurrent work spans LLM post-training (SFT/SDFT), retrieval-augmented generation, agentic reasoning, faithfulness and citation evaluation, and live clinical deployment — among othersAs an early-career scientist, you will grow fast — owning real features end-to-end, learning from a strong multi-disciplinary team, and meeting the actual medical professionals who will be using our systems🚩 Roles & ResponsibilitiesBuild, implement, and help design components of ‘Clinical Intelligence': medical foundation models, LLM- and retrieval-based AI systems that address real-world clinical problemsRapidly turn ideas into working prototypes and deployable PoCs, iterating quickly with the teamContribute to the core components of ‘Clinical Intelligence' — medical RAG, agentic reasoning pipelines, context management, LLM post-training (SFT/SDFT/RL), and evaluation of faithfulness, citation quality, and hallucinationWrite high-quality, well-tested code in our internal research codebases, following solid engineering practicesCollaborate with clinical partners (hospitals, medical doctors) and product teams to ship systems clinicians actually useRequirements🎯 Qualifications2+ years of industry/research experience in machine learning or deep learning, or a strong portfolio of built-and-shipped projects (new graduates with demonstrated building experience are welcome), with a preference for natural language processing, large language models, information retrieval, or medical AIHands-on proficiency in Python and modern ML frameworks (e.g., PyTorch, Hugging Face Transformers), with working knowledge of LLM, NLP, and retrieval techniques and the ability to build end-to-end (e.g., vLLM, RAG frameworks such as LlamaIndex or LangChain)Demonstrated track record of building and shipping working systems — side projects, open-source, internships, hackathons, or competitionsTaste for automation and engineering quality‘Bias towards action' — a strong builder instinct; you shipFast learner, comfortable with ambiguity and rapid iterationCollaborative team playerCapable of handling day-to-day business communication in English (written or verbal)Highly responsible, with an eye for detail and motivated to build high-quality, reliable solutions following current best practicesStrong motivation to work on medically impactful problems and contribute to advancing the standard of care through AI🏅 Preferred ExperiencesExperience working with clinical, biomedical, or otherwise high-stakes text data (ie. EMR)Hands-on experience with one or more of: LLM fine-tuning / post-training (SFT, DPO/RLHF), retrieval-augmented generation systems, agentic / tool-using LLM pipelines, or LLM evaluation frameworksExperience deploying a system into real-world production (cloud or on-premise)Experience with context engineering / context management for large-context LLM systems — structuring and orchestrating knowledge, tools, and state across an agent's context windowUp-to-date awareness of emerging trends and recent advances in LLM, agent, and retrieval researchContributions to open-source repositories in NLP, LLMs, RAG, or related areasAble to engage with customers (clinicians, hospital staff) to define problems and design systems📝How To ApplyCV (English or Korean, free format)Any other relevant material (English or Korean) *optionalLinks to a portfolio, GitHub, or shipped projects are strongly encouraged🏃‍♀️ Hiring ProcessDocument Screening → Introductory Interview → Technical Interview → Culture-fit Interview → OnboardingAfter the final interview, we may proceed with reference checks if needed🤝 Work Conditions and EnvironmentWork Type: Full-Time(3-month probation)Work Location : Lunit HQ (5F, 374, Gangnam-daero, Gangnam-gu, Seoul, 06241, Republic of Korea)Salary: After Negotiation🎸 ETCIf you misrepresent your experience or education or provide false or fraudulent information in or with your application, it may be grounds for cancellation of the employmentLunit is committed in providing the preferential processing to those eligible for employment protection (national merits and people with disabilities) relevant to related laws and regulationsBenefits🌻Benefits & PerksThe office is at a very convenient location, just a minute away from Gangnam Station Exit 3Meal Allowance is provided (up to 12,000 KRW per meal) when working at the officeLatest computer models, such as Macs and 4K monitors are provided and can be renewed every three yearsSeminar registration fees and book purchases are coveredRegular in-house AI and medical seminars are heldIn-house English lessons (aka Luniversal) is provided for English developmentAccess to high-quality AI learning resources & deep learning DevOps systemUp to 1.2 million KRW worth of benefits points can be claimed annuallyHoliday Allowances are provided in the form of gifts or vouchers for Korean National holidays, Seollal and ChuseokCongratulatory and Condolence allowances, along with paid time off are providedAnnual medical checkups and employee accident insurance are provided