Member of Technical Staff - Pre-Training Data
Reflection AISan Francisco, CAOn-Site
Needs a free Superjobs account. You'll finish on Reflection AI's Ashby page.
Skill score
Summary
The person designs and runs techniques for sourcing, curating, filtering and measuring large-scale web, text and code data for pre-training, and runs ablations to trace how data changes affect model behavior. It suits someone with strong Python and deep learning skills who understands data quality, statistics and data ethics.
- Function
- Data Science
- Estimated salary
- Est. $215k–323k/yr
- Estimated for a staff-level role, ~9 yrs experience (typical for the level), United States. Not published by the employer.
- Work mode and location
- On-Site, San Francisco, CA
- Employment type
- Full-time
Job description
Roles and responsibilities
- Design and implement techniques for sourcing, curating, and filtering large-scale web, text, and code data, with opportunities to extend into other modalities as our models evolve.
- Develop data quality metrics and analyses that measure coverage, diversity, and representativeness across sources.
- Join the pre-training research team to design and run data experiments, working with infrastructure teams to scale the supporting data systems efficiently and reproducibly
- Investigate and mitigate data risks, including privacy, safety, and licensing concerns, to support responsible and ethical data use.
- Evaluate dataset changes through ablations and scaling experiments, tracing their downstream effects on model learning and behavior.
- Publish and present research, and share code, datasets, and insights that move the open AI community forward.
Skills required
Top skills
- Python
- PyTorch or JAX
- Distributed workloads debugging
All skills (6)
- Large-scale data curation
- Data quality metrics
- Probability and statistics
- Multimodal data
- Data licensing and ethics
- Scaling experiments
Qualifications
- Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch or JAX), with experience debugging distributed workloads and writing code that scales.
- Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
- Clear written communication and the ability to explain complex technical ideas.
Nice to have
- A strong grasp of probability, statistics, and ML fundamentals, and the ability to tell real effects from noise and bugs in experimental data.
- Experience with curation, preprocessing, and analysis of large-scale text or code datasets; exposure to multimodal data is a plus.
- Background in data engineering, dataset construction, or large-scale web data processing for machine learning.
- Experience evaluating or improving training data quality, and knowledge of data ethics, safety, and licensing frameworks relevant to AI datasets.
- Contributions to open datasets, research publications, or data tooling.
- PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline; or equivalent industry research experience.
Other details
- Top-tier compensation: Salary and equity structured to recognize and retain our talent globally.
- Stock options: Everyone who joins and contributes to Reflection's success gets to share in the upside through stock options.
- Health & wellness: Comprehensive medical, dental, vision, and life, with an annual wellness allowance.
- Meals: Lunch and dinner are provided in the office daily.
- Life & family: 22 weeks paid parental leave for all new birthing and non-birthing parents, including adoptive and surrogate journeys.
- Vacation days: Unlimited paid time off in the U.S. and 30 days in the U.K.
- Sponsorship support: We sponsor visas to help exceptional talent join our team and support long-term immigration pathways where applicable.
- Team building: We have regular off-sites, happy hours, and team celebrations.
Company info
About Reflection AI
Reflection AI builds open-weight AI models and systems that give enterprises, governments, and sovereign entities control over their own intelligence. Its models are used for on-premises enterprise AI deployments, by U.S. National Laboratories through the Department of Energy's Genesis Mission, and in a planned sovereign AI cloud for enterprises and