Titlesby GlyphResearch titles
Scientists and research engineers who train, evaluate and study models.
Research EngineerWrites the code that makes research run: training runs, experiments, infrastructure.1.6% of HN posts, 12 mo · 161 open · $340k median · since Apr 2011Research ScientistDesigns and runs experiments to push what models can do; usually a PhD role.1.4% of HN posts, 12 mo · 122 open · $232k median · since Mar 2012Member of Technical StaffA deliberately flat title from Bell Labs, revived by AI labs for engineers and researchers alike.1.1% of HN posts, 12 mo · 206 open · $250k median · since Aug 2015Applied ScientistResearch methods pointed at a product problem; Amazon made the title mainstream.0.39% of HN posts, 12 mo · 26 open · $243k median · since Feb 2016AI TutorWrites and grades data that teaches a model a language or a subject; a model-training role, not teaching people.10 open · seen Jul 2020AI ScientistBuilds and trains new AI models for one field, such as biology or enterprise data, from research methods to the models a product relies on.3 open · $275k median · seen Sep 2026Machine Learning ScientistDesigns and trains machine learning models for a product problem, such as growth experiments, cancer diagnostics or search, and ships them.15 open · $193k median · seen May 2024AI TrainerTeaches an AI system by example: labels data such as photos of vehicle damage, or tunes a support agent's prompts, tone and guardrails.4 open · seen May 2026AI Safety ResearcherStudies how to make models behave as intended: alignment, interpretability, red teaming.0% of HN posts, 12 mo · 2 open · $153k median · since Mar 2024AI Red TeamerAttacks models before release: finds jailbreaks, harmful outputs and misuse, and turns them into fixes and evals.0% of HN posts, 12 mo · since Aug 2024Model Behavior DesignerSpecifies and tunes how a model talks and decides: its personality, refusals and defaults.0% of HN posts, 12 mo · new on job boards