AI Research Scientist, Multi-Agent AI Systems

3 weeks ago

San Francisco Bay Area, United States Stealth Startup Full time

Who We Are

We are a privately backed AI research team composed of scientists and engineers whose work has appeared in venues like Nature, Science, NeurIPS, ICLR, ICML, CVPR, ECCV, AAAI, IJCAI, ICRA, CoRL, AAMAS, and LoG. Our backgrounds span top research labs and academic centers (e.g., Stanford, UC Berkeley, MIT, CMU, Oxford, Tsinghua, IIT, Harvard, KAIST, Peking, Zhejiang, Shanghai Jiao Tong, ETH Zurich, Caltech). Together, we’ve achieved breakthroughs in multi-agent coordination, communication, and emergent behaviors with real-world impact.


By emphasizing thoughtful experimentation over brute force, we consistently reduce rework and produce tangible insights. Our members have improved negotiation efficiencies between agents and discovered architectures that simplify cooperative tasks. We focus on what truly matters—rigorous hypotheses, reproducible methods, and minimal wasted effort.


Responsibilities

You will design and conduct research on multi-agent systems that pushes the boundaries of how agents learn, reason, and collaborate. A deliberate, methodical approach drives your work, ensuring each experiment yields meaningful results.


You’ll collaborate with peers who value clarity, efficiency, and continual refinement. Your findings and perspectives will inform the entire team’s direction—whether that means revisiting how reward structures are defined or refining a data pipeline to eliminate unnecessary complexity.


Your Role & Impact

You will design and conduct research on multi-agent systems that pushes the boundaries of how agents learn, reason, and collaborate. A deliberate, methodical approach drives your work, ensuring each experiment yields meaningful results.


You’ll collaborate with peers who value clarity, efficiency, and continual refinement. Your findings and perspectives will inform the entire team’s direction—whether that means revisiting how reward structures are defined or refining a data pipeline to eliminate unnecessary complexity.


Qualifications

Technical Expertise:

  • Strong background evidenced through demonstrated expertise in multiple of the following areas, pushing at least one area forward at the frontiers of research or application: Representation learning, deep reinforcement learning, meta-learning, continual learning, unsupervised, semi, or self-supervised learning, GNNs, large language models, large multimodal models, multimodality, dataset, benchmark, evaluation and metric design and application, or robotics, ideally with exposure to multi-agent scenarios.
  • Proficiency in modern AI frameworks (e.g., PyTorch, TensorFlow, JAX), DevOps/hardware (e.g., GPU acceleration, HPC clusters, Docker/Kubernetes), and Visualization Tools and Algorithms (e.g., for Data, Models, Train/Test/Production changes)
  • Familiarity with relevant multi-agent libraries or frameworks (e.g., MetaGPT, AutoGen, LangGraph) or comparable tools that push cooperative, competitive, hybrid, centralized, and/or decentralized AI systems forward.

Research Mindset:

  • A history of conducting, publishing, and contributing to significant AI experiments. Citation count doesn't matter; ingenuity, experimental and theoretical rigor, and clearly motivated potential significance of work at time of presentation do.
  • Comfortable designing smaller, carefully scoped experiments that validate key insights rather than massive, unfocused runs.
  • Ability to spot workflow inefficiencies, propose improvements, and collaborate to make them happen.
  • You take the role of research scientist, engineer, and product manager because you want to own your work and have the ability to do so. This being said, we have multiple engineering groups to propel overall work, and we have no PMs that aren't qualified to do the research or engineering work they support so that we don't hamstring overall work.

Communication & Collaboration:

  • Skilled at presenting complex ideas in a clear, logical manner, both in writing and in discussions.
  • Eager to learn from and mentor colleagues, creating a cycle of mutual growth and shared discovery.
  • Thrive in an environment in which you're given the trust and agency to provide asynchronous updates when you are able to, you attend only the parts of meetings beneficial to you and save 10+ hours per 40 hours of your time (yes, we've conducted rigorous statistical hypothesis tests supporting this with lower than 0.05 p-value), and you are encouraged, even rewarded, for speaking up with kindness, logical reasoning, and overarching desire to maximize positive impact.


Your First Months Here

First Month:

  • Get hands-on with our internal architectures, reproducing a baseline experiment to internalize key methods.
  • Observe and reduce a redundant step in the workflow, immediately saving the team valuable time.

First Three Months:

  • Lead at least one experiment or project that explores new avenues of agent cooperation, potentially laying the foundation for a high-impact publication.
  • Mentor a junior team member, instilling the value of careful hypothesis testing and minimal rework.
  • Suggest evidence-based enhancements to our research pipeline, staying true to our focus on efficient, high-quality outputs.


Compensation & Environment

Compensation:

  • Base salary $300k–$500k, equity $1M-$4M + discretionary bonuses that collectively may exceed 100% base salary in a single calendar year
  • Total compensation is highly variable, but final offers result from a holistic review by a panel, and not just single decision-maker.

Culture:

  • $30k+ annual budget for self-development (conferences, courses, workshops).
  • An inclusive atmosphere that rewards process improvements—cutting down busywork so you can focus on impactful research.
  • Access to robust compute infrastructure and a flexible schedule so you can do your best work.

Visa Sponsorship:

  • We aim to secure sponsorship for all qualified candidates. In rare cases, legal constraints may prevent this; however, we retain dedicated world-class immigration counsel to address potential issues.


How to Apply

We read every CV or résumé with care. Please submit yours along with any relevant work (publications, code repositories, research statements). You’re also welcome to include a brief note about why you find this research important or anything else you want us to know—though this is entirely optional. Much can be gleaned from your written work and follow-up discussions.


If you’re interested in applying rigorous, well-designed multi-agent research to unlock entirely new forms of intelligence—while continually refining the process to keep it lean and efficient—let’s talk.

  • AI Agent Systems Researcher

    3 weeks ago

    San Francisco, California, United States Naptha AI Full time

    Job DescriptionNaptha AI is seeking an exceptional Research Scientist intern to contribute to advancing the state of AI agent systems. This internship offers hands-on research experience at the frontier of AI technology, backed by industry veterans and technical leaders.We're building the foundational infrastructure for the next wave of AI companies,...

  • Research Scientist

    7 days ago

    San Francisco, United States Naptha AI Full time

    Job DescriptionJob DescriptionResearch Scientist (AI Agent Behavior) | naptha.aiAbout this roleWe are seeking an exceptional Agent Behavior Scientist to study and optimize how AI agents interact, collaborate, and evolve within large-scale networks. This is a rare opportunity to shape the future of AI agent infrastructure at a massively ambitious scale,...

  • Research Scientist

    7 days ago

    San Francisco, United States Naptha AI Full time

    Job DescriptionJob DescriptionResearch Scientist (AI Agent Economies) | naptha.aiAbout this roleWe are seeking an exceptional Agent Economics Researcher to study and design incentive structures and economic systems for large-scale AI agent networks. This is a rare opportunity to shape the future of AI agent infrastructure at a massively ambitious scale,...

  • Research Scientist

    7 days ago

    San Francisco, United States Naptha AI Full time

    Job DescriptionJob DescriptionAI Research Scientist (Multi-agent Systems) | naptha.aiAbout the roleWe are seeking an exceptional AI Research Scientist to join Naptha AI at the ground floor, focusing on advancing the state of the art in multi-agent systems and agent interoperability. In this role, you will be responsible for researching and developing novel...

  • AI Agent Economies Researcher

    6 days ago

    San Francisco, California, United States Naptha AI Full time

    Job OverviewNaptha AI is a pioneering organization at the forefront of AI agent infrastructure development. We are seeking an exceptional Research Scientist to study and design incentive structures and economic systems for large-scale AI agent networks.About This RoleWe're building foundational infrastructure for the next wave of AI companies, enabling...

  • Multidisciplinary AI Research Expert

    7 days ago

    San Francisco, California, United States Naptha AI Full time

    About the RoleNaptha AI is seeking an exceptional AI Research Scientist to join our team at the ground floor, focusing on advancing the state of the art in multi-agent systems and agent interoperability.This role is critical in solving core technical challenges around agent communication protocols, task delegation, and distributed intelligence. You will work...

  • AI Agent Economy Specialist

    3 weeks ago

    San Francisco, California, United States Naptha AI Full time

    We are Naptha AI, a leading company in AI agent infrastructure, and we are looking for a talented Research Scientist to join our team. As a Research Scientist, you will play a key role in designing the economic foundations of future AI systems, including designing economic mechanisms for agent systems, game theory in multi-agent networks, and resource...

  • Research Internship in Multi-Agent Systems

    3 weeks ago

    San Francisco, California, United States Naptha AI Full time

    Research OpportunitiesNaptha AI is offering a unique opportunity for a Research Intern to contribute to advancing the state of AI agent systems. This internship provides hands-on research experience at the frontier of AI technology, backed by industry veterans and technical leaders.We're building the foundational infrastructure for the next wave of AI...

  • AI Research Scientist

    3 weeks ago

    San Francisco, California, United States Scale AI Full time

    Role Overview:">Scale AI is seeking a highly skilled AI Research Scientist/Research Engineer to join our Safety, Evaluations and Analysis Lab (SEAL) team. As a key member of this pioneering research effort, you will play a crucial role in developing robust evaluation products and tackling challenging research problems in evaluation and red teaming.">About...

  • Senior AI Behavior Researcher

    3 weeks ago

    San Francisco, California, United States Naptha AI Full time

    Job OverviewNaptha AI is seeking an exceptional American-dollar Estimated Salary: $200,000 - $250,000 per year Agent Behavior Scientist to study and optimize how AI agents interact, collaborate, and evolve within large-scale networks.About this RoleThis is a rare opportunity to shape the future of AI agent infrastructure at a massively ambitious scale,...

  • Stealth Startup | AI Research Scientist, Multi-Agent AI Systems

    2 weeks ago

    san francisco, United States Stealth Startup Full time

    Who We AreWe are a privately backed AI research team composed of scientists and engineers whose work has appeared in venues like Nature, Science, NeurIPS, ICLR, ICML, CVPR, ECCV, AAAI, IJCAI, ICRA, CoRL, AAMAS, and LoG. Our backgrounds span top research labs and academic centers (e.g., Stanford, UC Berkeley, MIT, CMU, Oxford, Tsinghua, IIT, Harvard, KAIST,...

  • Research Scientist

    7 days ago

    San Francisco, United States ZipRecruiter Full time

    Job DescriptionResearch Scientist (AI Agent Economies) | naptha.aiAbout this roleWe are seeking an exceptional Agent Economics Researcher to study and design incentive structures and economic systems for large-scale AI agent networks. This is a rare opportunity to shape the future of AI agent infrastructure at a massively ambitious scale, backed by industry...

  • AI Research Engineer

    6 days ago

    San Francisco, United States Perplexity AI Full time

    Perplexity is seeking experienced AI Research Engineers and Scientists to own the AI at the core of our answer engine. Your job is to provide a seamless experience between default and Pro search, while continuing to add new advanced agentic capabilities. Applicants should be well versed in all aspects of LLMs from prompting to training to evaluating in order...

  • Research Scientist

    2 months ago

    San Francisco, United States techire ai Full time

    Do you want to join a research team working on cutting-edge foundational multimodal models?You’ll be an experienced Researcher / Scientist who wants to shape the future of multimodal generative AI. You’ll need to have worked on complex AI models and have relevant research publications in either:Vision generation - video, image, 3D, LVMsAudio generation -...

  • Agent Systems Optimization Lead

    3 weeks ago

    San Francisco, California, United States Naptha AI Full time

    Job DescriptionNaptha AI seeks a skilled Agent Behavior Scientist to drive innovation in AI agent infrastructure. This role offers a competitive salary range of $200,000 - $250,000 per year.The successful candidate will be responsible for analyzing agent behavior patterns, designing experiments, developing measurement frameworks, creating agent behavior...

  • AI Research Engineer

    5 days ago

    San Francisco, United States Perplexity AI Full time

    Job DescriptionJob DescriptionPerplexity is seeking experienced AI Research Engineers and Scientists to own the AI at the core of our answer engine. Your job is to provide a seamless experience between default and Pro search, while continuing to add new advanced agentic capabilities. Applicants should be well versed in all aspects of LLMs from prompting to...

  • Research Scientist, AI Systems

    4 weeks ago

    San Francisco, California, United States RI Research Instruments GmbH Full time

    About the Position">Job Summary:We are seeking a highly motivated Research Scientist, AI Systems to join our team at RI Research Instruments GmbH. As a scientist, you will contribute to the development of new AI systems that benefit society as a whole.About the Team:Our team is dedicated to advancing the development of AI that is reliable, interpretable, and...

  • Data Scientist

    3 weeks ago

    San Francisco, California, United States Perplexity AI Full time

    We are seeking a skilled Data Scientist to join our team at Perplexity AI.OverviewAs an early member of the Data Science team, you will play a pivotal role in shaping the direction of our product and company. Our mission is to be the world's most knowledge-centric company by reshaping the future of search and technology.Role OverviewYou will have the chance...

  • AI Research and Development Engineer

    3 weeks ago

    San Francisco, California, United States Scale AI, Inc. Full time

    About Scale AI, Inc.We are accelerating the development of AI applications to make them accessible across every industry. Our mission is to transform how organizations build and deploy AI, powering the world's most advanced LLMs, generative models, and computer vision models.Our team is expanding to drive innovation in AI research and development. As a...

  • Staff Research Scientist

    3 days ago

    San Jose, California, United States Hume AI Inc Full time

    At Hume AI Inc, we're revolutionizing the field of AI by developing empathic systems that truly understand human behavior and preferences. Our innovative approach to AI development focuses on reinforcement learning from human expression (RLHE), which enables our models to learn from subtle cues like tone of voice, facial expressions, and more.About UsWe're a...