Data Scientist - Extensions
Fundamental · Remote · Europe · Israel (remote) ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Fully remote
- 5+ years experience
- Python + XGBoost
- Tabular ML focus
- Equity included
Join Fundamental's AI research lab to improve the predictive performance of NEXUS, the world's most powerful Large Tabular Model. You will tackle hard data science problems, collaborating with R&D and Engineering teams to ship production-grade solutions that transform how Fortune 100 companies make decisions.
The work spans diverse industries and prediction tasks, requiring deep expertise in traditional ML models and real-world tabular data challenges. You will design robust Python components, run rigorous experiments, and translate findings into tangible product capabilities.
What you'll do
- Research methods to improve NEXUS predictive performance across diverse datasets
- Design and implement robust, production-quality Python components
- Develop strategies to handle real-world enterprise data characteristics reliably
- Run rigorous experiments and design benchmarks to guide prioritization
- Solve structured data problems including classification, regression, and forecasting
- Collaborate with Engineering and Research teams to understand model behavior
- Validate approaches on customer datasets with Applied AI Engineers
- Contribute to technical documentation and internal best practices
What you'll need
- 5+ years of experience in data science or machine learning roles
- Strong Python skills with fluency in pandas, numpy, and scikit-learn
- Deep hands-on experience with XGBoost, LightGBM, CatBoost, or similar
- Solid understanding of tabular data challenges like class imbalance and missing values
- Strong experimental mindset for designing benchmarks and analyzing noisy results
- Ability to work autonomously and drive projects from idea to shipped output
Nice to have
- Familiarity with tabular foundation models like TabPFN or CARTE
- Competitive data science experience with top finishes in tabular competitions
- Background in finance, supply chain, healthcare, retail, or industrial domains
- Experience contributing to or designing internal ML libraries and shared tooling
- Familiarity with DuckDB, Polars, or modern in-process analytics engines
- Comfort reading ML research papers and translating findings into practice
Stack
- Python
- pandas
- numpy
- scikit-learn
- XGBoost
- LightGBM
- CatBoost
- DuckDB
- Polars
What you get
- Competitive compensation package with salary and equity
- Comprehensive health coverage for employees and dependents
- Paid parental leave for all new parents, including adoptive and surrogate
- Relocation support for employees moving to office locations
- Mission-driven culture valuing diversity of thought and ownership