Design and develop large-scale datasets to power generative AI models in multimodal domains (e.g., text, vision, speech), with a focus on synthetic data creation.
Build robust pipelines and tooling for data acquisition, cleaning, transformation, and quality assurance to support model training and evaluation.
Research, implement, and adapt cutting-edge techniques (e.g., fine-tuning, RLHF, data augmentation) to align generative models with domain-specific needs.
Curate and annotate datasets, ensuring diversity, representativeness, and compliance with responsible AI practices.
Evaluate open-source and research models, integrating best practices into data generation workflows.
Collaborate with engineering teams to ensure datasets and synthetic data pipelines are scalable, reliable, and production ready.
Develop metrics and benchmarking frameworks to assess data quality, model alignment, and downstream impact across modalities.
Partner cross-functionally with product, research, and infrastructure teams to drive innovation in data preparation and generative AI applications.

Qualification and Skills:

Bachelors or Master s in Computer Science, Data Science, AI/ML, or related field with 3+ years of industry experience.
Proficiency in Python and solid foundation in applied ML methods.
Proficiency with Pytorch, Torchvision, OpenCV, and similar, as well as building and deploying DNN models in production.
Experience building large-scale data pipelines for acquisition, cleaning, augmentation, and validation.
Ability to evaluate datasets for distribution, diversity, anomalies and fairness to assess overall quality and suitability for generative AI.
Experience with Computer Vision, NLP, Transformers, Large Language Models, Generative AI, optimizations around LLM training and serving. Experience with multimodal models a bonus.
Proven track record of delivering scalable, data-centric ML solutions.

Data Scientist 3