AI Reaserch Engineer

Description Sentra is on a mission to enable cloud-driven organizations to regain control and secure their full data lifecycle. We are a fast-growing start-up company, located in Tel Aviv (near a train station) with a very strong founding team. We're seeking a talented AI Research Engineer to join our Classification team in the R&D Data group. This role combines research, engineering and analytical thinking to make our AI-driven data classification accurate and reliable for our customers. We combine regex and rule-based classifiers with fine-tuned ML models (NER and text classification), along with great engineering and analytics, to deliver findings on millions of files and tables a day. Responsibilities • Research, design and ship new classification capabilities that expand coverage of sensitive data types, owning the end-to-end process from research and experiments through production deployment. • Evaluate and tune our classifiers (regex, dictionaries, fine-tuned NER and text classification models): build ground truth sets, measure precision and recall, find the root cause of false positives and false negatives, and fix them. • Build automated evaluation and benchmark workflows to compare classifier versions and catch quality regressions before they reach customers. • Design, build, and maintain data pipelines and dashboards to track classification quality and product performance, and deliver actionable insights to stakeholders at all levels. • Work with customers' real-world data problems: support PoCs and customers with deep classification analysis, and turn their feedback into product improvements. • Work from within the R&D Data team, take part in the discussions revolving around Sentra's AI research objectives, priorities and solutions. Requirements • 3+ years of hands-on experience in data analysis, data science or ML engineering in a B2B SaaS company • Proficiency with Python and SQL; experience with Spark or similar technologies • Strong regular expressions skills: write, test and tune patterns, and explain their false positive behavior • Experience designing experiments and measuring model quality (precision, recall, error analysis) • Experience building and maintaining data pipelines, ETL processes, and workflow orchestration tools (Airflow, Dagster, or similar) • Degree in a quantitative field such as Computer Science, Engineering, Mathematics, Statistics • Understanding of software development best practices (Git, CI/CD, testing) • Can-do attitude. Problem solving skills, self-starter, proactive mindset • Advantage: • • Previous experience in the cybersecurity domain (DSPM, DLP, data classification) • Experience with cloud environments (AWS, Azure, GCP) • Experience fine-tuning NLP models (for example NER or text classification) or using LLMs in daily work