Interested in this Research Scientist role at Amazon Web Services?
Apply Now →Skills & Technologies
About This Role
DESCRIPTION
Do you want to build the ML models that power every compliance verdict customers and auditors rely on? We are building a new Governance, Risk, and Compliance (GRC) service at AWS where every action — human or agent — is recorded, assessed, and provably governed, continuously and at any scale.
As an Applied Scientist on this team, you will own the Natural Language Inference (NLI) model at the center of our compliance engine: fine-tuning, calibrating, and monitoring it in production. You will also design the science behind framework decomposition, evidence auto-linking, and a new agent-context system that tackles retrieval, grounding, and hallucination measurement for AI agents. Your work will directly shape the accuracy and trustworthiness of every customer-visible result we produce.
Key job responsibilities
- Own the production NLI model end-to-end: fine-tune, promote, calibrate thresholds, and monitor for drift to ensure accurate compliance verdicts across frameworks.
- Design and implement framework decomposition pipelines that translate compliance controls into NLI hypotheses, including enrichment and per-framework threshold calibration.
- Build and improve the evidence auto-linking system using ranking, embedding, and reranking techniques, tuning precision-recall tradeoffs and addressing OCR quality challenges.
- Drive the quality flywheel by establishing labeling standards, designing offline and online evaluation methods, orchestrating retraining cycles, and ensuring privacy-safe data handling.
- Research and develop the agent-context capability — retrieval and representation of lifecycle signals, cross-framework knowledge, memory, grounding, and hallucination measurement — collaborating with engineering pods to bring these models into production.
A day in the life
You will split your time between collaborative and heads-down work. On the collaborative side, you will partner with several pods of software development engineers in an Agile cadence — aligning on data collection needs, sharing experiment results, and integrating models into production services. On the heads-down side, you will dig into model training runs, analyze evaluation metrics, and iterate on threshold tuning until the numbers meet the bar. Your work is roadmap-driven, giving you the space to plan experiments methodically while still responding to new framework onboarding requests as they arise.
About the team
Our team builds the AI and ML systems behind AWS Governance, Risk, & Compliance service, where science is not auxiliary — it is the core. Every customer-visible compliance result flows through models our scientists create and maintain. We are investing in next-generation capabilities including observability and governance for AI agents, and the multi-quarter Agent Context initiative is an organization-defining program that will need technical leaders.
Growth here is tangible. You will own the full model lifecycle from training through production monitoring, develop rare expertise at the intersection of ML and compliance, and have a clear path to expand your scope into technical leadership as the service scales.BASIC QUALIFICATIONS
- 2+ years of building models for business application experience
- 2+ years of programming in Java, C++, Python or related language experience
- 2+ years of hands-on predictive modeling and large data analysis experience
PREFERRED QUALIFICATIONS
- PhD in computer science, machine learning, engineering, or related fields
- Experience in written and verbal communication with the ability to present complex technical information in a clear and concise manner to executives and non-technical leaders
- Experience with one of the following areas: machine learning technologies, Reinforcement Learning, Deep Learning, Computer Vision, Natural Language Processing (NLP) or related applications
- Experience in any of the following areas: algorithms and data structures, parsing, numerical optimization, data mining, parallel and distributed computing, high-performance computing
- Extensive experience applying theoretical models in an applied environment.
- Experience with building agentic systems using LLMs and familiar with common agentic framework such as Strands Agent, LangGraph, etc.
- Solid knowledge on LLM techniques and experience with LLM post-training.
Amazon is an equal opportunity employer and does not discriminate on the basis of protected veteran status, disability, or other legally protected status.
Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, including support for the interview or onboarding process, please visit https://amazon.jobs/content/en/how-we-hire/accommodations for more information. If the country/region you’re applying in isn’t listed, please contact your Recruiting Partner.
The base salary range for this position is listed below. Your Amazon package will include sign-on payments and restricted stock units (RSUs). Final compensation will be determined based on factors including experience, qualifications, and location. Amazon also offers comprehensive benefits including health insurance (medical, dental, vision, prescription, Basic Life & AD&D insurance and option for Supplemental life plans, EAP, Mental Health Support, Medical Advice Line, Flexible Spending Accounts, Adoption and Surrogacy Reimbursement coverage), 401(k) matching, paid time off, and parental leave. Learn more about our benefits at https://amazon.jobs/en/benefits.
USA, VA, Arlington - 142,800.00 - 193,200.00 USD annually
USA, WA, Seattle - 142,800.00 - 193,200.00 USD annually
Salary Context
This $142K-$193K range is below the median for Research Scientist roles in our dataset (median: $196K across 91 roles with salary data).
Role Details
About This Role
Research Scientists push the boundaries of what AI can do. They design experiments, develop novel architectures, publish papers, and translate research breakthroughs into production capabilities. This is where the fundamental advances happen, from attention mechanisms to diffusion models to reasoning chains.
The work is intellectually demanding and often ambiguous. You might spend months on an approach that doesn't pan out. The best research scientists combine deep mathematical intuition with engineering pragmatism. They know when to go deep on theory and when to run experiments. They read papers voraciously and can spot incremental contributions from genuine breakthroughs.
Across the 10,419 AI roles we're tracking, Research Scientist positions make up 2% of the market. At Amazon Web Services, this role fits into their broader AI and engineering organization.
Research Scientist roles are concentrated at major AI labs (OpenAI, Anthropic, Google DeepMind, Meta FAIR) and well-funded AI startups. The competition is intense. PhD is effectively required for most positions, and publication track record matters. Compensation is among the highest in AI, reflecting both the scarcity of talent and the strategic importance of research breakthroughs.
What the Work Looks Like
A typical week includes: reading and discussing recent papers with your team, designing and running experiments on multi-GPU clusters, analyzing results and iterating on hypotheses, writing up findings for internal review or publication, and collaborating with engineering teams to productionize promising results. The ratio of thinking to coding is higher than in engineering roles.
Research Scientist roles are concentrated at major AI labs (OpenAI, Anthropic, Google DeepMind, Meta FAIR) and well-funded AI startups. The competition is intense. PhD is effectively required for most positions, and publication track record matters. Compensation is among the highest in AI, reflecting both the scarcity of talent and the strategic importance of research breakthroughs.
Skills Required
PhD strongly preferred for most roles. Deep expertise in a specific area (NLP, computer vision, reinforcement learning, multimodal) is expected. PyTorch is the standard. Publication track record matters. Strong mathematical foundations in linear algebra, probability, optimization, and information theory are assumed.
Beyond the fundamentals, companies value experience with large-scale distributed training, novel architecture design, and the ability to bridge theory and practice. Understanding of current frontier topics (reasoning, multimodal, long-context, alignment) is essential. Code quality matters more than many researchers expect. Labs want researchers who can implement their ideas cleanly.
Strong research postings specify the research area, mention the team you'd join, and describe the problems they're working on. They often list recent publications from the team. Vague 'AI research' postings without specifics usually mean the company wants to sound impressive but doesn't have a real research agenda.
Compensation Benchmarks
Research Scientist roles pay a median of $223,400 based on 917 positions with disclosed compensation. Mid-level AI roles across all categories have a median of $193,200. This role's midpoint ($168K) sits 25% below the category median. Disclosed range: $142K to $193K.
Across all AI roles, the market median is $214,834. Top-quartile compensation starts at $266,000. The 90th percentile reaches $321,730. For comparison, the highest-paying categories include AI Safety ($274,500) and Research Engineer ($252,000). By seniority level: Entry: $114,400; Mid: $193,200; Senior: $230,000; Director: $272,100; VP: $235,000.
Amazon Web Services AI Hiring
Amazon Web Services has 54 open AI roles right now. They're hiring across AI/ML Engineer, AI Product Manager, Research Scientist, AI Engineering Manager. Positions span New York, NY, US, Cupertino, CA, US, Brooklyn, NY, US. Compensation range: $142K - $303K.
Location Context
Across all AI roles, 6% (580 positions) offer remote work, while 9,776 require on-site attendance. Top AI hiring metros: New York (3,785 roles, $222,480 median); San Francisco (3,262 roles, $261,300 median); Los Angeles (1,896 roles, $210,000 median).
Career Path
Common paths into Research Scientist roles include PhD Student, Research Engineer, Postdoc.
From here, career progression typically leads toward Research Lead, Distinguished Scientist, VP of Research.
The PhD is the entry point for most paths. Choose your advisor and research area carefully since they'll define your first industry position. Publish consistently, contribute to open-source projects in your area, and build relationships at conferences. Industry research offers better compensation and compute resources than academia, but the pressure to show product impact is real.
What to Expect in Interviews
Research interviews are multi-stage: a research talk (present your best paper), technical deep-dives on your methodology, and often a 'research proposal' exercise where you design an experiment to test a hypothesis. Coding rounds test implementation ability alongside theoretical knowledge. Be prepared to implement a paper from scratch and discuss the design choices the authors made. Strong candidates can critique papers constructively and identify gaps in experimental methodology.
When evaluating opportunities: Strong research postings specify the research area, mention the team you'd join, and describe the problems they're working on. They often list recent publications from the team. Vague 'AI research' postings without specifics usually mean the company wants to sound impressive but doesn't have a real research agenda.
AI Hiring Overview
The AI job market has 10,419 open positions tracked in our dataset. By seniority: 117 entry-level, 4,975 mid-level, 3,840 senior, and 1,487 leadership roles (Director, VP, C-Level). Remote roles make up 6% of the market (580 positions). The remaining 9,776 roles require on-site or hybrid attendance.
The market median for AI roles is $214,834. Top-quartile compensation starts at $266,000. The 90th percentile reaches $321,730. Highest-paying categories: AI Safety ($274,500 median, 90 roles); Research Engineer ($252,000 median, 580 roles); AI Engineering Manager ($250,811 median, 72 roles).
Research Scientist roles are concentrated at major AI labs (OpenAI, Anthropic, Google DeepMind, Meta FAIR) and well-funded AI startups. The competition is intense. PhD is effectively required for most positions, and publication track record matters. Compensation is among the highest in AI, reflecting both the scarcity of talent and the strategic importance of research breakthroughs.
The AI Job Market Today
The AI job market spans 10,419 open positions across 16 role categories. The largest categories by volume: AI/ML Engineer (7,849), AI Product Manager (516), AI Software Engineer (513). These three account for the majority of open positions, though smaller categories often have higher per-role compensation because of specialized skill requirements.
The seniority mix tells a story about where AI teams are in their maturity. Entry-level roles (117) are outnumbered by mid-level (4,975) and senior (3,840) positions, reflecting that most companies are past the 'build a team from scratch' phase and need experienced engineers who can ship production systems. Leadership roles (Director, VP, C-Level) total 1,487 positions, representing the bottleneck between technical execution and organizational strategy.
Remote work availability sits at 6% of all AI roles (580 positions), with 9,776 requiring on-site or hybrid attendance. The remote share has stabilized after the post-pandemic correction. Senior and specialized roles (Research Scientist, ML Architect) are more likely to be remote-eligible than entry-level positions, partly because experienced hires have more negotiating power and partly because these roles require less hands-on mentorship.
AI compensation is structured in clear tiers. The market median sits at $214,834. Top-quartile roles start at $266,000, and the 90th percentile reaches $321,730. These figures include base salary with disclosed compensation. Total compensation (including equity, bonuses, and sign-on) runs 20-40% higher at companies that offer those components.
Category matters for compensation. AI Safety roles lead at $274,500 median, while Prompt Engineer roles sit at $140,000. The spread between highest and lowest-paying categories reflects the premium on specialized technical skills versus broader analytical roles.
The most in-demand skills across all AI postings: Python (2,350 postings), Aws (1,173 postings), Azure (1,030 postings), Rag (986 postings), Gcp (714 postings), Pytorch (649 postings), Kubernetes (583 postings), Claude (581 postings). Python dominates, appearing in the vast majority of role descriptions regardless of category. Cloud platform experience (AWS, GCP, Azure) is the second most common requirement. The newer entrants to the top skills list (RAG, vector databases, LLM APIs) reflect the shift from traditional ML toward generative AI applications.
Frequently Asked Questions
Get Weekly AI Career Intelligence
Salary data, skills demand, and market signals from 16,000+ AI job postings. Every Monday.