Data Scientist – AI/ML Modeling

This role designs, validates, productizes custom machine learning and deep learning models for laboratory solutions. The Data Scientist will convert scientific workflow challenges into rigorous modeling problems, develop fit-for-purpose algorithms, establish defensible evaluation methods and partner with product, software, application science, quality and field teams to deliver reliable AI-enabled capabilities. The emphasis is on practical model development for laboratory data and scientific workflows—not on LLM or foundation-model specialization. Priority use cases may include chromatographic and mass spectrometry data analysis, image analysis, anomaly detection, model-assisted review and workflow intelligence.

• →Convert AI product requirements, customer workflow pain points and scientific use cases into clear ML problem statements, model objectives, evaluation metrics and experimental plans.
• →Design, train, validate and optimize custom ML/DL models for laboratory productivity use cases such as chromatography, mass spectrometry, image analysis, anomaly detection and model-assisted review.
• →Select and adapt suitable architectures for scientific data, including convolutional, segmentation, temporal, graph-based, multimodal, or physics-informed approaches where appropriate.
• →Build reproducible modeling workflows covering data curation, labeling strategy, ground-truth definition, feature engineering, quality checks, training, evaluation, and documentation.
• →Work with application scientists and domain experts to design experiments and generate high-quality datasets for model development and benchmarking.
• →Define performance metrics, statistical validation approaches, acceptance criteria, and evidence packages suitable for scientific and product decision-making.
• →Translate prototypes into product-ready capabilities by collaborating with software engineering on model interfaces, inference workflows, APIs, performance constraints, deployment patterns, and lifecycle needs.
• →Ensure model outputs are scientifically valid, explainable, reproducible and aligned with customer workflow expectations.
• →Document assumptions, datasets, experiments, results, limitations, risks, and validation evidence to support quality review, responsible AI practices and maintainability.
• →Monitor applied AI/ML research and selectively evaluate methods that can improve laboratory data analysis, automation and workflow intelligence with clear practical deployment value.

Qualifications

• Master’s degree or Ph.D. preferred in data science, machine learning, statistics, computer science, applied mathematics, bioinformatics, chemometrics, analytical chemistry, life sciences, or a related quantitative discipline.
• Minimum 5 years of relevant industry or applied research experience in machine learning, deep learning, data science, or scientific computing; advanced degree research experience may be considered.
• Strong hands-on Python experience with commonly used ML libraries and scientific computing tools such as PyTorch, TensorFlow, scikit-learn, NumPy, Pandas, or equivalent frameworks.
• Solid understanding of ML/DL fundamentals, including model selection, training strategy, validation design, optimization, evaluation metrics, and error analysis.
• Ability to structure ambiguous scientific or product problems into testable hypotheses, data plans, modeling approaches, experiments, and success criteria.
• Strong quantitative foundation in statistics, applied mathematics, optimization, signal processing, time-series analysis, computer vision, chemometrics, or adjacent methods for scientific data.
• Familiarity with reproducible ML practices, including experiment tracking, data traceability, model versioning, technical documentation and collaboration with software engineering teams.
• Fluent English and strong communication skills for technical documentation, experiment reviews, and collaboration with global stakeholders.

• Applied AI/ML experience in laboratory, life science, analytical instrumentation, scientific workflow, or customer-facing product environments.
• Domain exposure to chromatography, mass spectrometry, peak integration, peak detection, baseline correction, spectral analysis, scientific imaging, or instrument-generated signal data.
• Experience tailoring advanced model architectures to noisy, limited, heterogeneous, imbalanced, or instrument-dependent scientific datasets.
• Practical experience improving model robustness, explainability, uncertainty handling, or performance consistency across real-world operating conditions.
• Experience transitioning models beyond proof of concept, including inference design, model serving, APIs, containerization, CI/CD integration, monitoring, or lifecycle management.
• Exposure to LLMs, foundation models, or AI-assisted software development tools is a plus, but not a core requirement.