DATA SCIENTIST - GenAI & Cybersecurity

Job description

We are looking for a Data Scientist to develop and improve Generative AI models for our product, including security data understanding, reasoning, and incident summarization.

You will focus on model development, training data, fine-tuning, and evaluation, working with AI Engineers and security specialists to turn research and experiments into practical product capabilities.

Key responsibilities

• Research, develop, and adapt GenAI models for cybersecurity use cases.

• Build and curate training datasets from security data, incident reports, and expert annotations, ensuring quality and appropriate train-test separation.

• Experiment with model architectures, training approaches, and fine-tuning techniques to improve domain-specific performance.

• Explore synthetic data generation and domain adaptation to address limited or specialized training data.

• Build evaluation frameworks to measure factual correctness, reasoning quality, evidence grounding, and task completion.

• Analyze model failures and improve data, training configurations, and model behavior through reproducible experiments.

• Explore model optimization techniques to balance quality, latency, and computational cost.

• Partner with AI Engineers to bring validated models into production and incorporate feedback into further development

Qualifications

• 3-4 years of experience in data science, applied ML, or AI research, including hands-on work developing or fine-tuning GenAI models.

• Strong Python skills and experience with PyTorch, Hugging Face, or similar tools.

• Solid understanding of deep learning, transformer architectures, and generative modeling fundamentals.

• Practical experience preparing training data and building reproducible training and evaluation workflows.

• Familiarity with GPU-based experimentation and techniques for efficient training.

• Ability to design experiments, diagnose model issues, and explain results and limitations clearly.

• A strong passion for cybersecurity and curiosity about applying GenAI to security operations.

Nice to have :

• Previous experience applying AI in the cybersecurity domain.

• Experience with instruction tuning, LoRA/PEFT, preference optimization, or domain-specific continued pretraining.

• Experience with synthetic data generation, model distillation, or quantization.

• Experience with multimodal models or training models for tool use and multi-step tasks.

• Experience taking research prototypes into production with engineering teams.