AI Research Engineer Jobs at Startups

Luma is seeking a Research Scientist/Engineer to build the systems that enable reinforcement learning (RL) at a frontier scale. This role involves coupling policy optimization with large fleets of inference workers, agentic environments, and reward/verification systems to translate model behavior into learning signals. RL is crucial for Luma's models to evolve from capable to useful. This is a full-loop systems problem encompassing training, rollout generation, environment execution, and reward computation across thousands of GPUs, requiring speed, stability, and correctness. The ideal candidate has experience with post-training LLMs using RL, building environments and verifiers, and debugging large-scale asynchronous rollout pipelines.

Salary not listed

Invisible Technologies is seeking a Research Engineer to join its research organization, focusing on reinforcement learning, evaluation methodology, benchmarks, and RL environments for frontier labs and enterprise clients. This role involves defining how frontier models are measured, with direct influence over the methodology and its implementation. You will translate research questions into practical scoring frameworks, evaluation architectures, or environment designs, and then build the production systems to run them. This is a hands-on role for someone who wants to design the approach and write the code to prove it out, with reporting lines varying by experience level.

Salary not listed

HUD is building reliable, fair, and open infrastructure for AI data, aiming to make data valuable for creators and trustworthy for AI labs. As a Research Engineer on the Robotics team, you will develop datasets and evaluations crucial for training and assessing embodied AI systems. This role involves translating research needs into data specifications, creating methods for data validation, and conducting experiments to understand how data characteristics impact model performance. You will collaborate with internal research and engineering teams, and potentially with data vendors and buyers, to enhance robotics data offerings.

$100k - $230k

Percepta

Percepta is transforming critical institutions with applied AI, embedding with industry-leading customers to drive AI transformation. We bring together forward-deployed expertise in engineering, product, and research, along with our in-house toolkit for rapidly deploying agentic workflows and strategic partnerships. As a Research Scientist - Reinforcement Learning, you will work at the intersection of RL research and real-world deployment, advancing decision-making capabilities for critical industries. You will collaborate closely with Embedded Product Managers and engineers to ensure our solutions transform how companies operate.

$200k - $400k

Handshake is seeking a Member of Technical Staff, Evals to help define how frontier AI systems are measured, understood, and improved. This is a high-ownership role for researchers who build, partnering with AI researchers, domain experts, and customers to develop new benchmarks, reward and verifier systems, agent-evaluation methodologies, and data-quality techniques. Early members of the team will have significant influence over technical direction, operating culture, and the open-source software, benchmarks, and research products built. The company prioritizes demonstrated research capability, technical judgment, and a builder's mindset over specific titles, degrees, or career paths.

$200k - $350k

SecureBio is seeking a Senior Research Scientist to join their AI Biosecurity Assessments team. In this role, you will evaluate frontier AI models for dual-use biological risks before their public release. This involves running technical assessment pipelines, analyzing model behavior, and producing reports that inform relevant audiences about identified risks. Your work will contribute to setting industry safety standards and influencing policy bodies like the NIST US AI Safety Consortium and the European AI Office. You will collaborate with AI firms, internal research teams, and policy makers to ensure the safe development and deployment of advanced AI technologies.

$235k - $300k

About HUD HUD's mission is to build reliable, fair and open infrastructure for AI data. We want data to be valuable for the people who create it and trustworthy for the labs that train on it. Our team is a quickly growing group of researchers, engineers and operators building the economy that shapes what AI will become. Backed by $16M from top VCs and YC (W25), our marketplace and platform are used by startups, Fortune 500 companies and frontier labs. About the role We’re looking for a Research Engineer for our Robotics team to develop the datasets and evals that make robotics data useful for training and evaluating embodied AI systems. You’ll translate open-ended research needs into data specifications, build methods to validate data, and run experiments to understand which data and structures improve model performance. Responsibilities - Research the data needs of robot learning and physical AI systems, and turn them into concrete dataset and evaluation specifications - Define data schemas, annotations, ground truth, and quality standards across robotics data types - Design collection and review protocols that external data providers can execute reliably - Build tools and validation workflows to audit datasets, identify quality issues, and give providers actionable feedback - Run experiments and analyze model behavior to understand how data quality, coverage, and structure affect performance - Work with HUD’s research and engineering teams, and when needed with data vendors and buyers, to improve robotics data offerings Experience You may be a good fit if you have: - Experience in robotics, robot learning, embodied AI, or closely related multimodal research - Proficiency in Python and experience building data processing, analysis, or evaluation tools - Experience turning research questions into dataset specifications, experiments, and measurable quality criteria - Strong understanding of what makes robotics data useful for training or evaluation—and where it can be misleading - Attention to detail and the ability to spot subtle errors, coverage gaps, and failure modes in complex data - Experience building research tools or pipelines without a fully prescribed roadmap Strong candidates may also have: - Worked with robot trajectories, demonstrations, video, sensor data, simulation, or other multimodal robotics datasets - Experience with imitation learning, reinforcement learning, or vision-language-action models - Worked in unstructured problem spaces and take ownership from early research through production deployment - Early-stage startup experience and strong communication skills for collaboration across teams and time zones We prioritize technical aptitude and learning potential over years of experience. Motivated candidates are encouraged to apply even if they don't meet all criteria. Team & company details - Team Size: ~25 people currently, mostly full-time in-person, but some remote. - Our team: Our team includes 4 International Olympiad medalists (IOI, ILO, IPhO), serial AI startup founders, and researchers with publications at ICLR, NeurIPS, etc. - Company stage: We have 8 figures in funding and are scaling profitably and quickly to meet very strong demand. Logistics - Employment: Full-time. - Location: We have offices in San Francisco or Singapore but are open to remote candidates who can work hours that 70-80% overlap with either San Francisco or Singapore time zones. - Visa Sponsorship: We provide support for relocation and visas for strong full-time candidates to the US or Singapore. - Timeline: Applications are rolling. The process is 2 technical interviews and a 2-3 day work trial. What we offer - Competitive compensation - 100% covered top-of-the-line medical, dental, and vision (US and Singapore employees) - Lunch and dinner for in-office employees - Company-wide holiday break (Christmas Eve to New Year’s Day) on top of PTO and paid holidays - Other perks including an Equinox membership, 401k, and commuter benefits (US employees) and a health/wellness stipend (Singapore employees) - Unlimited* access to tokens for ChatGPT, Claude Code, Cursor, etc. *By unlimited, we mean no one on our token usage leaderboard has ever hit a limit. So we have no idea what the limit is. - Two top-floor offices in San Francisco’s Union Square and Singapore’s Raffles Place - Annual travel budget to visit either office Compensation Actual offers are adjusted for experience and location, but our base salary bands are - San Francisco (and other major US cities): $135,000 - $230,000 - Singapore: $100,000 - $175,000 - Rest of world: $100,000 - $175,000 Due to high volume, we may not actively respond to every application, but feel free to contact us at [email protected] or elsewhere if we missed your application!

$100k - $230k

Tavus

Tavus is building the human layer of AI, aiming to make human-AI interaction as natural as face-to-face communication. We pioneer research in multimodal AI for modeling human communication (language, audio, video) and generating audio-visual avatar behavior, powering applications from text-to-video avatars to real-time conversational video. We are a Series B company backed by top investors, seeking individuals who thrive in startup environments, can prioritize independently, and are willing to take calculated risks to help pave the path forward.

Salary not listed

HUD is building reliable, fair, and open infrastructure for AI data, aiming to make data valuable for creators and trustworthy for AI trainers. We are a growing team of researchers, engineers, and operators backed by significant VC funding. We are seeking a Research Manager to lead research focused on enhancing HUD's agent training data and evaluations for frontier models. This role involves guiding research engineers through complex projects, translating findings into scalable methods, and staying technically engaged while ensuring rigorous experimentation and impactful results. The ideal candidate will possess a deep understanding of model training dynamics, including how signals, data, and evaluation design influence outcomes.

$120k - $295k

Eventual is building data curation for Physical AI, processing petabyte-scale video and sensor data to extract signals about how hands and objects interact, temporal events, and clip reliability for training frontier models. We are a small, well-backed team from leading tech companies, aiming to revolutionize how Physical AI models are trained by providing reliable, curated datasets derived from vast amounts of video footage. This role is based in our SF Mission District office, with a 4-day work week. As Technical Lead, Multimodal Research, you will drive the technical vision for Eventual's understanding of video data. You will own the strategy for selecting, prototyping, and deploying models and representations at scale, ensuring they meet rigorous evaluation standards and cost-efficiency targets. This is a senior individual contributor role focused on setting research direction and making key architectural decisions, while remaining hands-on with research and experimentation.

$150k - $300k

Cartesia

Cartesia is pioneering AI architectures that learn and interact with the world like humans. We are seeking a Research Engineer, Data Infrastructure to build the datasets and systems essential for pretraining our foundational models. This role involves developing performant, scalable infrastructure for acquiring, processing, and curating massive datasets, and collaborating closely with research teams to optimize data composition. Your contributions will directly influence the capabilities and quality of our cutting-edge models.

$200k - $350k

Together AI

As a Research Intern in the Model Shaping team, you will contribute to advanced post-training methods, explore new techniques for efficient neural network training, and work on robust evaluation of foundation model capabilities. The Model Shaping team focuses on tailoring open foundation models for downstream applications, developing services for model selection and improvement, and creating new methods for efficient model training and evaluation. This role offers the opportunity to work on cutting-edge research, potentially leading to publications in top-tier conferences.

$58k - $70k

Together AI

As a Research Intern in the Model Shaping team, you will contribute to advanced post-training methods, explore new techniques for efficient neural network training, and work on robust evaluation of foundation model capabilities. The Model Shaping team focuses on tailoring open foundation models for downstream applications, developing services for model selection and improvement, and creating new methods for efficient model training and evaluation. This role offers the opportunity to work on cutting-edge research, potentially leading to publications in top-tier conferences.

$58k - $70k

Together AI

The Inference Research team is dedicated to building the next generation of efficient, scalable, and reliable serving systems for large foundation models, advancing open and transparent AI. This role operates at the intersection of model architectures, high-performance systems engineering, and hardware optimization, focusing on co-designing software, algorithms, and models to reduce the cost and latency of AI systems. As a research intern, you will explore distributed inference, compiler-aware optimization, and novel inference-time computation strategies, co-designing and implementing cross-layer optimizations across models, systems, and hardware, with a focus on KV cache design and large-scale serving architectures. Projects aim to unlock unprecedented performance and scale for foundation models, enabling faster serving, larger model deployment, and robust evaluation under realistic serving workloads.

Salary not listed

Together AI

The Inference Research team is dedicated to building the next generation of efficient, scalable, and reliable serving systems for large foundation models, advancing open and transparent AI. This role operates at the intersection of model architectures, high-performance systems engineering, and hardware optimization, focusing on co-designing software, algorithms, and models to reduce the cost and latency of AI systems. As a research intern, you will explore distributed inference, compiler-aware optimization, and novel inference-time computation strategies, co-designing and implementing cross-layer optimizations across models, systems, and hardware.

Salary not listed

Together AI

The Agents team focuses on building, aligning, and scaling advanced AI systems capable of handling complex, multi-step tasks and workflows across text and speech, particularly in agentic and scientific domains. This role sits at the intersection of agent capabilities, human-computer interaction, and infrastructure, exploring areas like post-training methods for agentic behavior and evaluation frameworks for open-ended tasks. As a research intern, you will contribute to cutting-edge agentic AI challenges, addressing issues of alignment, reliability, and scalability. Projects may involve developing new training techniques for self-learning and long-horizon reasoning, curating datasets for scientific and agentic tasks, analyzing agentic behavior failure modes, or building scalable agent operation infrastructure.

Salary not listed

Together AI

The Agents team investigates how to build, align, and scale frontier AI systems capable of complex, multi-step tasks across text and speech, focusing on agentic and scientific domains. This role sits at the intersection of agent capabilities, human-computer interaction, and infrastructure, exploring areas like post-training methods for agentic behavior and evaluation frameworks for open-ended tasks. As a research intern, you will tackle challenges in alignment, reliability, and scalability, potentially working on new training recipes for self-learning, curating datasets for non-deterministic tasks, studying agentic behavior failure modes, or building large-scale agent operation infrastructure.

Salary not listed

Together AI

Together AI's Model Shaping team focuses on tailoring open foundation models for downstream applications, building services that help ML developers select and improve models with domain-specific data. The team also develops efficient model training and evaluation methods, drawing from machine learning, NLP, and ML systems. As a Research Engineer, you will develop a platform for users to customize open-source models with their own data, working across training and inference stacks. You will build and enhance services for Fine-Tuning, Reinforcement Learning, and Evaluation, ensuring a smooth transition from post-training to production serving and optimizing inference engines for RL training. Collaboration with product, research, and engineering teams is key to maintaining API reliability and performance, ultimately contributing to the foundational layer of the open-source AI ecosystem.

$200k - $290k

As a Senior Research Engineer at Escape, you will join a team of talented security researchers focused on innovating and translating cutting-edge research into impactful production solutions. You will play a key role in developing agentic scanning systems that autonomously explore applications, designing and implementing a robust security detection engine for various vulnerabilities, and integrating AI, specifically LLMs, into scanning workflows to enhance coverage and accuracy. Your work will also involve building and optimizing distributed infrastructure for large-scale concurrent scans, ensuring performance and reliability, and contributing to core platform architecture decisions.

Salary not listed

Cursor

As a Research Engineer on the Synth Data team at SpaceXAI, you will invent and scale the synthetic data that teaches our models new capabilities. You will own the full loop: spotting a capability the model is missing, designing a data generation strategy, building reliable pipelines and verifiers, and validating that the resulting data actually moves the model. You will form hypotheses about what data will help, build the systems to produce it at scale, run ablations, and read the resulting traces and training signals to understand why it worked (or didn’t). Your datasets will go directly into frontier training runs and shape the development of Grok models.

Salary not listed

Genmo is a research lab focused on developing advanced video world models to understand, simulate, and interact with the physical world, aiming to advance physical intelligence and enable robots to learn and act. We are seeking an exceptional Software Engineer to join our research team in pushing the boundaries of visual generative AI. In this role, you will collaborate with experienced researchers to develop and enhance generative models, ensuring their safe and effective deployment. This is an opportunity to contribute to cutting-edge AI research in a fast-paced startup environment.

Salary not listed

LanceDB is seeking an AI Research Engineer to collaborate with their research team on open-ended research projects. The role focuses on end-to-end training flows across various AI domains, demonstrating how LanceDB can accelerate research. This is a unique opportunity to pursue research interests as an engineer, significantly impacting product development and raising awareness of LanceDB's capabilities.

$150k - $255k