ARTIFICIAL INTELLIGENCE (AI) SCIENTIST

Job Responsibilities:

1. Responsible for understanding and breaking down data-related tasks for the company’s SkyReels video large model, including assessing video content features, data quality, task complexity, business objectives, and technical feasibility, and formulating corresponding data strategies and rule systems.

2. Lead the data production workflow for video and image-related large model projects, including designing annotation guidelines, decision boundaries and quality control standards, and driving the large-scale development and optimisation of the data pipeline covering data collection, screening, annotation, quality inspection and evaluation.

3. Work closely with algorithm, annotation and evaluation teams to improve model performance through data-driven iteration, and establish task-level feedback loops between data production, model training and model evaluation.

Requirements:

1. PhD or Master’s degree in Artificial Intelligence, Computer Science, Electrical/Electronic Engineering, Mathematics, Multimedia, or a related field, or equivalent practical research experience.
   
2. At least 4 years of relevant R&D experience in video understanding, video generation, multimodal algorithms, or related fields.
   
3. Experience with large models, video-related data algorithms, or large-scale AI implementation is preferred.
   
4. Strong coding skills in Python and at least one other programming language such as C++, Java, or Go, with a strong foundation in algorithms and data structures.
   
5. Strong understanding of deep learning theories and experience with deep learning frameworks such as PyTorch or TensorFlow.
   
6. Familiarity with video generation foundation models, diffusion models, or related video generation architectures.
   
7. Experience with multimodal model development, such as CLIP, BLIP, or other vision-language models, and ability to design training strategies for cross-modal alignment.
   
8. Familiarity with distributed computing, big data processing, or model training/inference frameworks such as Ray, Flink, Spark, PyTorch, or vLLM.
   
9. Strong ability to break down complex problems, with good team management skills, result orientation, and end-to-end planning capabilities.
   
10. Proficiency in Mandarin is preferred, as the role requires frequent collaboration with Mandarin-speaking stakeholders.