Research Sr. Data Scientist – NLP and LLM Data
Pubblicato il 18-09-2026 - Equixly in Verona
About Equixly
Equixly is an Italian cybersecurity company founded to revolutionise API security through proprietary artificial intelligence and machine learning algorithms. Our platform automates the API Security Testing process, enabling companies to identify deep vulnerabilities rapidly, scalably, and continuously.
On a strong growth trajectory, we are building out our AI Team with talented engineers and scientists who want to work on real, production-grade LLM systems and high-complexity technical challenges at the cutting edge of applied AI.
The role
We are looking for a Research Senior Data Scientist to join the AI Team and contribute to the development, analysis, curation, and evaluation of the data used to train and improve Equixly’s AI systems.
This role sits at the intersection of data science, machine learning, NLP, LLM evaluation, and applied AI. You will work on data quality, evaluation methodology, statistical analysis, and data-centric improvements for models used in real cybersecurity products.
This is a high-impact position with direct influence on the quality,
reliability, and performance of the models that power our platform. You will report to the Head of AI.
What you will do
Your key responsibilities will include:
- Design, analyse, curate, and improve datasets for LLMs, NLP systems, classifiers, and agentic AI workflows.
- Support the preparation of high-quality data for model training, evaluation, benchmarking, and continuous improvement.
- Define data quality standards, labelling principles, evaluation protocols, and reproducible analysis practices.
- Analyse model behaviour, errors, edge cases, and failure modes to identify opportunities for improvement.
- Collaborate with AI engineers and researchers to connect data decisions with model performance, robustness, and production behaviour.
- Document datasets, experiments, evaluation methodology, and key findings clearly and rigorously.
- Stay aligned with the state of the art in data-centric A