Large Language Model (LLM) Algorithm Engineer

**Job Overview** We are looking for talents (fresh graduates or professionals with 1–3 years of experience) to join our LLM team. You will take part in the R\&D of LLM applications and Agent systems, and practice end-to-end engineering methodologies — covering prompt engineering, model post-training, and turning conceptual ideas into deliverable online solutions. **Key Responsibilitie** 1. sEngage in the design, development and optimization of LLM applications and Agent systems, including prompt engineering, Tool Use / MCP, multi-agent orchestration, etc 2. .Carry out model post-training tasks: data construction, training and evaluation for alignment technologies like SFT and RFT 3. .Conduct model fine-tuning experiments based on llama-factory, ms-swift and other post-training frameworks, and build reusable training and evaluation pipelines 4. .Leverage AI coding tools such as Claude Code and Codex for development, and explore optimal engineering practices for human-AI collaboration 5. .Follow the latest advances in LLM fields (research papers, open-source projects, new base models), evaluate relevant technologies and deploy them into business scenarios **Job Requiremen** 1. tsMaster’s degree or above in Computer Science, Software Engineering, Artificial Intelligence, Mathematics or other related discipline 2. s.Open to fresh graduates and talents with 1–3 years of work experience; candidates with prior LLM/algorithm experience are preferre 3. d.Hong Kong local resident (holding HK Permanent / Non-permanent Identity Card), or holder of valid Hong Kong work visa (IANG, TTPS Top Talent Pass Scheme, QMAS Quality Migrant Admission Scheme, etc. 4. ).Genuine passion for AI and solid understanding of development concepts, including Prompt Engineering and Model Post-Trainin 5. g.Sound knowledge of core LLM fundamentals (Transformer, attention mechanism, pre-training \& post-training paradigms) and mainstream application architectures (RAG, Agent, Function Calling 6. ).Skilled in at least one mainstream post-training framework such as llama-factory or ms-swift; practical hands-on fine-tuning experience is a plu 7. s.Frequent user of AI coding tools (Claude Code, Codex, Cursor, etc.) with clear insights on real-world usag 8. e.Strong Python programming capability, familiar with SQL, with standardized coding habits and good engineering literac **Preferred Qualificati** 1. onsPublished LLM research papers at top conferences (ACL / EMNLP / NeurIPS / ICLR, etc.) or active contributions to open-source projec 2. ts.Practical development experience with Agent frameworks (LangChain, Claude Agent SDK, etc.) or full RLHF training pipelin 3. es.0-to-1 full lifecycle delivery experience of LLM products and projects (for candidates with 1–3 years of experience); prizes in Kaggle, ACM and other algorithm competitio 4. ns.Personal AI portfolio (GitHub repositories, technical blogs, online demos, etc.) available for presentation during in interviews.