Research & Science
Advancing AI capabilities, applying AI to scientific discovery, and providing domain expertise to improve AI systems. Covers AI research, ML research, safety/alignment research, applied science, computational biology/chemistry/physics, quantitative research, AI tutoring and RLHF evaluation, and domain expertise for AI training and evaluation.
Roles
The canonical roles within Research & Science.
Research Scientist
Research scientists in these roles formulate and execute high-impact research problems spanning multimodal AI, video understanding, generative modeling, and autonomous systems, often balancing fundamental innovation with product integration. They distinguish themselves by combining exceptional experimental judgment with the ability to identify and frame novel problems where existing benchmarks are insufficient, rather than simply executing well-defined research directions. These scientists typically work within interdisciplinary research teams at major AI labs and well-funded startups, collaborating closely with ML engineers to translate advances into production systems while maintaining the rigor needed for publication at top-tier venues.
AI Tutor & Domain Expert
Domain experts apply specialized knowledge to strengthen AI systems through hands-on work in data annotation, model evaluation, and training refinement. These professionals leverage deep expertise in specific fields—from psychology and audio engineering to business operations and customer support—to create high-quality training datasets and provide critical feedback that shapes how AI models behave. They work closely with technical teams to translate real-world problem-solving into actionable data that improves model reasoning, accuracy, and domain-specific performance. What distinguishes this work is the direct expertise requirement; practitioners must combine genuine mastery in their subject area with the ability to decompose complex problems into trainable signals for AI systems. These roles typically sit within dedicated human data or training teams at AI companies, collaborating with machine learning engineers and product teams to ensure models learn nuanced, accurate representations of their domains.
Research Engineer
Research Engineers at these organizations work across the full stack—from implementing cutting-edge algorithms and optimizing models for specialized hardware, to building scalable infrastructure that translates research prototypes into production systems. They combine deep machine learning expertise with strong software engineering skills, often bridging gaps between research scientists and infrastructure teams to accelerate progress on frontier AI problems like inference optimization, reinforcement learning for robotics and reasoning, multimodal generation, and agentic systems. These roles typically sit within research teams that collaborate closely with product and infrastructure groups, requiring engineers to balance scientific rigor with practical engineering constraints while contributing to publications and deployments that advance the field.
Member of Technical Staff
Members of Technical Staff at AI labs drive core breakthroughs in model development by owning critical junctures in the training pipeline—from data strategy and synthetic generation through pre-training, mid-training, and post-training optimization. They combine deep research insight with engineering rigor to inject capabilities across reasoning, coding, mathematics, and multimodal understanding, translating empirical findings into measurable improvements that shape what models can fundamentally do. These roles sit at the intersection of research and systems engineering within small, talent-dense teams, where they work cross-functionally to ensure that raw model intelligence becomes aligned, safe, and deployable at scale—balancing theoretical innovation with pragmatic delivery against real-world constraints.
Applied ML Scientist
Applied ML Scientists design and optimize machine learning systems that solve concrete business or scientific problems, moving beyond theoretical research to ship models in production environments. They work at the intersection of modeling and systems engineering, combining cutting-edge techniques like fine-tuning, reinforcement learning, and synthetic data generation with practical constraints around latency, cost, and real-world data distribution. These roles typically sit within dedicated applied research or product teams at AI-native companies, collaborating closely with engineers and domain experts to translate customer requirements or product challenges into effective training pipelines and evaluation frameworks.
Research Management
Research managers in this role oversee teams developing AI solutions for complex scientific and technical challenges, from drug discovery and autonomous systems to model evaluation and safety research. They balance hands-on technical leadership with strategic planning, setting research directions and priorities while mentoring scientists and engineers through exploratory work. What distinguishes these leaders is their ability to translate frontier AI research into measurable outcomes—whether that's evaluating model capabilities, optimizing machine learning pipelines, building new evaluation frameworks, or steering teams toward products that solve real customer problems. They typically operate within specialized research functions nested within larger product or engineering organizations, working cross-functionally to ensure research breakthroughs integrate into platforms and services that matter.
Simulation Engineer
Simulation Engineers at AI companies build and operate the simulation infrastructure that supports training, validation, and design across two broadly distinct domains. The first is robotics and autonomous-systems simulation—physics solvers, sensor simulators, and high-fidelity virtual environments for training and validating robots, autonomous vehicles, and other embodied systems. The second is scientific and engineering simulation—finite-element analysis, computational fluid dynamics, atomistic and molecular simulation—used by AI-for-science and TechBio companies to validate predictions and generate training data. The two paths share methodological backbone (numerical methods, validation against experiment, automation at scale) but draw on different deep technical foundations. These engineers typically sit within dedicated simulation, research, or platform teams, collaborating with ML researchers, domain scientists, or robotics integrators depending on the application.
Physical & Life Scientist
Physical and life scientists in AI companies design and execute experiments across biology, chemistry, and physics to accelerate drug discovery and therapeutic development. These roles span from wet lab work—running immunological assays, synthetic chemistry, and analytical separations—to computational approaches like molecular dynamics simulations and pharmacometric modeling. What distinguishes these scientists is their direct integration with AI-driven platforms: they validate predictions from machine learning models, generate training data for foundation models in biology and chemistry, conduct safety evaluations of AI systems in scientific domains, and translate computational designs into experimental reality. They typically sit within discovery and development teams at TechBio and AI companies, collaborating closely with computational researchers, engineers, and medicinal chemists to bridge the gap between digital prediction and physical validation.
Recent Jobs
The latest Research & Science openings across the AI industry.