Member of Technical Staff, Applied Research
$150k - $350k • San Francisco • FullTime
Posted 1mo ago
Job Location
San Francisco
Tech Stack
Remote Work Policy
On-site
Employment Type
FullTime
Categories
AI Engineer
About the job
Sieve is seeking a Member of Technical Staff, Applied Research to join their team. In this role, you will train and evaluate multimodal models, focusing on video generation and audiovisual understanding, to understand how data influences model capabilities. You will own the research process end-to-end, from identifying model weaknesses and forming hypotheses to building experiments and evaluating results. This involves fine-tuning models, developing reproducible training and evaluation pipelines, and conducting controlled experiments on data quality and supervision. The ideal candidate enjoys bridging research and engineering, implementing methods, debugging training runs, and ensuring findings are reliable and impactful for data sourcing, curation, and usage.
Responsibilities
- Train and post-train models for video generation and multimodal understanding.
- Design controlled experiments to measure the impact of data selection, mixtures, and supervision on model capabilities.
- Build evaluations to identify specific model weaknesses using quantitative metrics and human judgment.
- Develop reliable training infrastructure, including distributed training, efficient data loading, checkpointing, and experiment tracking.
- Translate research findings into improvements for data curation pipelines and products.
- Collaborate with internal and partner research and engineering teams to define problems and communicate results.
Requirements
- 2+ years of experience in machine learning research or engineering.
- Hands-on experience training or fine-tuning deep learning models.
- Strong Python and PyTorch skills, including implementing, debugging, and modifying model training code.
- Experience designing experiments, establishing baselines, and critically evaluating results.
- Familiarity with modern generative or multimodal architectures (e.g., diffusion models, transformers).
- Comfort working with large datasets and diagnosing training bottlenecks, instability, and data quality issues.
- Ability to translate ambiguous research questions into concrete experiments and maintainable systems.
- Strong communication skills for explaining findings, tradeoffs, and uncertainty.
Benefits
- 401k
- Full Health Insurance
- Breakfast, Lunch, and Dinner covered
- Choice of snacks
- Ubers covered for travel home