Applied Scientist

We’re hiring an Applied Scientist to lead applied research engagements and turn scientific ideas into partner‑ready proofs of value while strengthening Sigma Nova’s reusable methods, tooling, and research agenda.

This is a hands‑on scientific role at the intersection of research, product thinking, and partner delivery: you will frame problems, run experiments, build demonstrators, and help translate results into decisions and deployable assets.

What you’ll work on

1) Frame applied research opportunities

• Translate a partner or internal problem into testable hypotheses and measurable success criteria.

• Assess data availability/quality/heterogeneity/sensitivity, and clarify data rights and constraints.

• Propose a realistic plan: state‑of‑the‑art, baseline, experimentation, demonstrator, validation, and (when relevant) transfer to production.

2) Design and execute rigorous scientific work

• Run targeted literature reviews and select the most relevant approaches.

• Develop baselines, experimental protocols, benchmarks, and ablations that are reproducible.

• Adapt advanced methods (including foundation models) to complex scientific/industrial data.

• Quantify performance, uncertainty, robustness, and limitations; perform error analysis.

3) Turn research into proofs of value

• Build fast demonstrators to test hypotheses with real constraints.

• Create reusable assets: model components, adaptation techniques, evaluation protocols, and comparison tooling.

• Translate results into clear, actionable recommendations for technical and non‑technical stakeholders.

• Connect benchmark metrics to real outcomes: uncertainty reduction, robustness gains, operational impact, and R&D acceleration.

• When relevant, contribute to publications, patents, academic collaborations, or tech transfer.

Preferred experience

What we’re looking for

• PhD in ML/AI/statistics/applied math/signal processing/computational physics (or MSc + exceptional applied research track record).

• Strong experience designing and evaluating ML/deep learning models.

• Ability to read literature, reproduce/adapt methods, and run rigorous experiments.

• Excellent Python skills and a deep learning framework (e.g., PyTorch), plus solid engineering/reproducibility practices.

• Comfort working with imperfect, scarce, heterogeneous, multimodal, or sensitive data.

• Strong communication and synthesis skills with scientific, technical, and business audiences.

Recruitment process

• Recruitment prescreen (30-45min)

• Scientific deep dive (remote-45min)

• Half-day of scientific interviews (Architecture - Coding - Research talk) + Culture fit

• References call