Senior Data Scientist

Company Description:

Harmoniz - The Art of Seed Innovation is a global agtech company specializing in breeding greenhouse seeds. The company is focused on developing innovative, great-tasting, and virus-resistant tomatoes, expanding into pepper and cucumber seed development.

Main Purpose of the Role:

The Senior Data Scientist will work closely with the core technical and domain experts on the hands-on development and evaluation of predictive models across the full spectrum of approaches — starting from traditional statistical and algorithmic baselines, progressing through classical machine learning methods as the core of the modeling effort, and exploring advanced deep learning architectures (CNNs and Transformers) as experimental options where they provide clear value. Prior experience with genomics-driven prediction, inheritance, and similarity modeling is a strong advantage.

Responsibilities:

• Work hands-on on the development, evaluation, and continuous improvement of predictive models across the full spectrum of approaches — from traditional statistical and algorithmic baselines, through classical machine learning methods as the core modeling layer, and up to experimental deep learning architectures (CNNs and Transformers) where they provide clear value.
• Research and implement algorithms for trait prediction, inheritance modeling, pedigree-based relationships, and similarity scoring across breeding lines, with genomics-driven approaches as a key growth direction.
• Build robust experimentation and benchmarking workflows, including reproducible comparisons between baseline algorithms, ML models, and advanced model variants, using appropriate evaluation metrics such as accuracy, precision/recall, ROC-AUC, calibration, regression error measures, and task-specific success criteria.
• Process and analyze very large-scale biological and breeding datasets, working with terabytes of genomic and phenotypic data while ensuring correctness, scalability, and reproducibility.
• Develop efficient computation pipelines and high-quality code, balancing model performance, runtime, infrastructure utilization, and cost constraints.
• Collaborate closely with core technical and domain experts to translate research outputs into practical, usable modeling capabilities that support real breeding and product decision-making.

Contribute actively within an agile R&D environment, aligning work with sprint progress, iterative milestones, and occasional deadline-driven execution

Qualifications

• 7+ years of experience in applied machine learning, data science, or algorithmic modeling, with strong hands-on ownership of predictive systems in production or applied research settings.
• MSc/PhD (2nd degree and above) in a relevant field such as Computer Science, Data Science, Bioinformatics, Computational Biology, Statistics, or equivalent — or proven comparable industry experience.
• Deep knowledge of algorithmics, statistical learning, and modern machine learning foundations, with demonstrated ability to implement complex predictive and optimization logic beyond off-the-shelf solutions.
• Proven experience developing and tuning classical machine learning models (regression, classification, ensemble methods, similarity-based approaches), including strong baseline construction and benchmarking discipline.
• Experience with deep learning frameworks such as TensorFlow (preferred) and/or PyTorch, and familiarity with CNN and Transformer architectures as advanced modeling tools when appropriate.
• Prior experience with genomics, computational biology, or genetic trait modeling is a strong advantage, particularly in the context of inheritance, similarity, and breeding prediction problems.
• Experience working with large-scale datasets and distributed compute environments, with the ability to design workflows that scale to terabytes of data.
• Experience applying computer vision or vision-based models to evaluate phenotypic traits is an advantage, especially in agriculture context.
• Strong software engineering practices, including clean research code, Git-based workflows, reproducible pipelines, and iterative experimentation discipline.
• Strong teamwork and collaboration skills, with the ability to work closely with a small group of core technical and domain experts in a fast-moving R&D environment.
• Familiarity with Databricks environments is an advantage (optional), particularly for large-scale biological computation and model experimentation.
• Fluent communication in Hebrew and English