What You’ll Do
1. Design efficient technical solutions leveraging scenario-driven data within the ecosystem. Lead application-layer optimization of large models, including model construction, distillation, fine-tuning, Agent/MCP frameworks, RAG (Retrieval-Augmented Generation), CoT (Chain-of-Thought), data synthesis, multimodal reasoning, long-text modeling, infrastructure architecture, automated deployment, third-party integration, and toolchain consolidation, as well as model tuning and performance enhancement for vertical-specific scenarios.
2. Keep close track of AI advancements in both academia and industry, identifying potential application scenarios and innovative value within the ecosystem.
What We’re Looking For
1. Bachelor’s degree or above in Computer Science, Mathematics, Physics, or related fields, with a strong mathematical foundation and at least 2 years of relevant work experience.
2. Publications in top-tier conferences such as **NeurIPS**, **ACL**, **ICML**, or **EMNLP** are preferred.
3. Hands-on experience in large model application development; familiarity with Agent frameworks, RAG enhancement, model alignment, and related optimization techniques is highly desirable.
4. Continuously follows cutting-edge AI technologies, with strong teamwork, communication, and collaboration skills; able to quickly adapt technical solutions to real-world scenarios and drive practical implementation.
5. Deep understanding of fundamental principles and common algorithms in Natural Language Processing, Machine Translation, and Computer Vision.
6. Strong grasp of self-attention and multi-head attention mechanisms, including positional encoding and layer normalization. Solid understanding of pretraining–fine-tuning paradigms and their application in large-scale models. Familiar with architectures and training methods of **BERT**, **GPT**, and similar language models.
7. Strong engineering skills, familiar with core data structures, operating systems, and programming concepts. Proficient in SQL, Python, PyTorch, Hugging Face, OpenCV, and experienced with deep learning frameworks such as TensorFlow and PyTorch.
8. Strong learning ability, innovative mindset, ownership mentality, excellent communication skills, and a collaborative team-oriented approach.
