Research Scientist

What you’ll do

1. 01Track leading model teams’ needs and movements across post-training, agent training, eval, RL environments, and synthetic data.
2. 02Build an iteration loop from training feedback to data strategy, grounded in model evals, customer feedback, and error analysis.
3. 03Independently own a research or data direction — problem definition, literature review, experiment design, data construction, evaluation methods, and retrospectives.
4. 04Systematically analyze how models fail at real creative tasks: reading the brief, aesthetic judgment, visual and spatial reasoning, tool use, multi-step iteration, and delivery quality.
5. 05Design benchmarks, evals, rubrics, quality standards, and training-data plans that genuinely serve capability gains.
6. 06Work with creative experts and the algorithm, data-engineering, and delivery teams to decompose expert judgment and real creative workflows into learnable, evaluable task units.

What we look for

1. 01Master’s or above in AI, CS, statistics, math, or a related field.
2. 02Experience with LLMs, NLP, multimodal, agents, model eval, post-training, data construction, or training-data analysis.
3. 03Understanding of SFT, RLHF / RLAIF, preference data, reward models, eval, agent training, and synthetic data.
4. 04Solid research training — paper reading, experiment design, data analysis, error attribution, and technical writing.
5. 05Strong coding ability — Python and common ML / data tools; able to build model calls, data processing, eval scripts, and experiment analysis.
6. 06Strong ownership — able to break fuzzy problems into clear goals, experimental paths, and deliverables.

Nice to have

• Conference papers, high-quality arXiv / workshop work, open-source projects, or research internships.
• Experience at a top AI lab, big-tech model team, research institution, or strong startup.
• Experience with post-training, benchmarks, eval, multimodal, code, image or 3D generation evaluation, or high-quality dataset construction.
• Your own views on capability boundaries, synthetic data, agent training, RL environments, or expert data.