Research Product Manager at Thinking Machines | San Francisco

At Thinking Machines Lab, we strive to empower humanity by advancing collaborative general intelligence. Our vision is to create a future where everyone can access the knowledge and tools necessary to harness AI for their specific needs and aspirations.Our team comprises scientists, engineers, and innovators who have developed some of the most widely utilize…

At Thinking Machines Lab, we are on a mission to empower humanity by advancing collaborative general intelligence. Our vision is to create a future where everyone has access to the knowledge and tools necessary to harness AI for their unique needs and objectives.We are a diverse team of scientists, engineers, and builders responsible for developing some of the most influential AI products on the market, such as ChatGPT and Character.ai. Our contributions extend to open-weight models like Mistral and popular open-source projects including PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the RoleWe are seeking talented engineers to join our team and develop the libraries and tools that will accelerate research efforts at Thinking Machines. You will take charge of our internal infrastructure—creating evaluation libraries, reinforcement learning training libraries, and experiment tracking platforms—while building systems that enhance research velocity over time.This position emphasizes collaboration. You will work closely with researchers to identify bottlenecks and pain points, ensuring that they trust your systems to function seamlessly and find them enjoyable to use.What You'll DoDesign, build, and manage research infrastructure, including evaluation frameworks, RL training systems, experiment tracking platforms, visualization tools, and shared utilities.Develop high-throughput, scalable pipelines for distributed evaluation, reward modeling, and multimodal assessment.Establish systems for reproducibility, traceability, and robust quality control across research experiments and model training runs, implementing effective monitoring and observability.Collaborate directly with researchers to identify bottlenecks and unlock new capabilities, managing research tools like a product manager by proactively seeking feedback and tracking adoption.Work alongside infrastructure, data, and product teams to integrate tools across the technical stack.

Thinking Machines Lab brings together scientists, engineers, and innovators behind widely recognized AI products such as ChatGPT and Character.ai, as well as open-source frameworks like PyTorch, OpenAI Gym, Fairseq, and Segment Anything. The team is driven by a mission to enhance humanity through collaborative general intelligence, aiming for a future where AI adapts to individual needs and goals. Tinker, the lab’s fine-tuning API, empowers researchers and developers to customize advanced AI models for their own use cases. Tinker manages the infrastructure, allowing users to train open-weight models with their chosen datasets, algorithms, and objectives. As Tinker grows its user base and features, the team is expanding to better support the community. Role overview The Forward Deployed Engineer acts as the main point of contact for a broad range of clients, from solo developers to large organizations. This role identifies customer challenges and requirements, then translates those insights into actionable product improvements. Both customer interaction and product development responsibilities are central to this position. What you will do Triage and resolve customer issues across the full stack, including analyzing logs, reproducing failures, and tracing job executions. Develop tools, integrations, and automation to address recurring problems and speed up user support. Create and update clear documentation and practical guides based on real user experiences and implementations. Work closely with research and infrastructure teams to turn customer feedback into prioritized engineering tasks. Help shape Tinker’s product roadmap by sharing insights from daily customer interactions.

Thinking Machines Lab aims to advance collaborative general intelligence, making AI accessible and adaptable for individuals and organizations. The team brings together scientists, engineers, and innovators behind well-known AI solutions, including ChatGPT, Character.ai, Mistral, and open-source projects like PyTorch, OpenAI Gym, Fairseq, and Segment Anything. Tinker, the lab’s fine-tuning API, helps researchers and developers customize AI models using their own data and algorithms. By handling the infrastructure, Tinker allows users to focus on training and deploying models that suit their needs. With a growing customer base and expanding features, the team is looking for a Software Engineer, Platform to support Tinker's continued development. Role overview This position centers on building and maintaining the core platform systems that power Tinker. The engineer will manage billing and usage metering, permissions and access control, organizational structures, data exports, audit logging, and the administrative tools that tie these systems together. Collaboration with product and legal teams is essential, as changes to features, pricing, and enterprise agreements will involve this role. What you will do Design the authorization layer for all products, including RBAC, API key scoping, organizational hierarchies, and permission boundaries. Oversee billing infrastructure, covering usage metering, plan management, payment processing, invoicing, and revenue recognition support. Develop and improve models for organizations and teams, such as seat management, SSO/SAML, workspace isolation, and invitation flows. Implement data export and deletion processes that align with enterprise standards and data residency requirements. Create audit logging systems to track user actions and decisions. This role is based in San Francisco.

At Thinking Machines Lab, our ambition is to enhance human potential by advancing collaborative general intelligence. We envision a future where individuals have the tools and knowledge to harness AI for their distinct requirements and aspirations.Our team comprises dedicated scientists, engineers, and innovators who have contributed to some of the most renowned AI products, including ChatGPT and Character.ai, along with open-weight models like Mistral, and influential open-source projects such as PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the RoleWe are seeking an Infrastructure Research Engineer to architect, optimize, and sustain the computational frameworks that facilitate large-scale language model training. You will create high-performance machine learning kernels (e.g., CUDA, CuTe, Triton), enable effective low-precision arithmetic operations, and enhance the distributed computing infrastructure essential for training expansive models.This position is ideal for an engineer who thrives in close collaboration with hardware and research disciplines. You will partner with researchers and systems architects to merge algorithmic design with hardware efficiency. Your responsibilities will include prototyping new kernel implementations, evaluating performance across various hardware generations, and helping to establish the numerical and parallelism strategies crucial for scaling next-generation AI systems.Note: This is an evergreen role that remains open continuously for expressions of interest. We receive numerous applications, and there may not always be an immediate opportunity that aligns with your qualifications. However, we encourage you to apply, as we regularly assess applications and will reach out as new positions become available. You are also welcome to reapply after gaining additional experience, but please refrain from applying more than once every six months. Additionally, you may notice postings for specific roles catering to particular projects or team needs. In such cases, you are encouraged to apply directly alongside this evergreen listing.What You’ll DoDesign and develop custom ML kernels (e.g., CUDA, CuTe, Triton) for key LLM operations such as attention, matrix multiplication, gating, and normalization, optimized for contemporary GPU and accelerator architectures.Conceptualize compute primitives aimed at alleviating memory bandwidth bottlenecks and enhancing kernel compute efficiency.Collaborate with research teams to synchronize kernel-level optimizations with model architecture and algorithmic objectives.Create and maintain a library of reusable kernels and performance benchmarks that serve as the foundation for internal model training.Contribute to the stability and scalability of our infrastructure, ensuring it meets the growing demands of AI development.

At Thinking Machines Lab, we are on a mission to enhance humanity through the advancement of collaborative general intelligence. Our vision is to create a future where everyone has the opportunity to leverage AI tailored to their individual needs and aspirations.Our team comprises scientists, engineers, and innovators who have developed some of the most renowned AI products in the industry, such as ChatGPT, Character.ai, as well as open-weight models like Mistral and popular open-source projects including PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the RoleWe are seeking an Executive Business Partner to provide vital support to several technical leaders from our San Francisco office. Your role will be crucial in ensuring our team remains focused and organized by managing personal logistics and handling tasks that may otherwise be overlooked.This position is unique, requiring creativity and flexibility to adapt to various work styles and the dynamic challenges of a fast-paced startup environment. You will enjoy significant autonomy in decision-making without extensive supervision.What You’ll DoManage calendars, schedule meetings, and coordinate travel for 3-4 technical leaders.Act as the primary liaison between your supported leaders and other departments within the company.Assist with recruiting coordination efforts.Monitor projects and commitments to ensure nothing is overlooked.

At Thinking Machines Lab, our mission is to empower humanity by advancing collaborative general intelligence. We envision a future where everyone has access to the knowledge and tools necessary to leverage AI for their unique goals.Our team consists of scientists, engineers, and builders who have developed some of the most utilized AI products, such as ChatGPT and Character.ai, alongside open-weight models like Mistral and popular open-source projects including PyTorch, OpenAI Gym, Fairseq, and Segment Anything.HR Business PartnerThe HR Business Partner role is essential in empowering our team to thrive as we scale. You will be pivotal in coaching our leaders and designing people systems that align with our mission.As the HR Business Partner, you will facilitate leadership coaching and the design of performance management systems that foster growth and collaboration. You will support managers in enhancing team dynamics and personal development while building a scalable people infrastructure that includes performance feedback systems, compensation structures, and career frameworks.What You’ll DoProvide coaching to managers by observing their leadership styles, identifying strengths and areas for growth, and promoting continuous improvement.Advise leadership on organizational strategies, including team structure, succession planning, and strategic people decisions that influence our operational effectiveness.Develop compensation frameworks that attract top-tier machine learning talent while ensuring alignment with our core values and principles.Create career progression frameworks tailored for a research environment where growth often transcends traditional management roles and where contributions such as mentorship and expertise are valued.Establish feedback and evaluation mechanisms that prioritize personal improvement over mere assessment.

Thinking Machines Lab brings together scientists, engineers, and innovators who have contributed to well-known AI products such as ChatGPT, Character.ai, and open-weight models like Mistral. The team’s open-source projects include PyTorch, OpenAI Gym, Fairseq, and Segment Anything. Their mission centers on advancing collaborative general intelligence and making AI tools accessible for a wide range of users and goals. The Tinker platform offers a fine-tuning API that lets researchers and developers tailor advanced AI models to their needs. By handling the underlying infrastructure, Tinker enables users to train open-weight models with custom data, algorithms, and objectives. As demand grows, the team is adding new features and supporting an expanding community. Role overview The Full Stack Software Engineer will play a key part in building and maintaining the products and services that Tinker users depend on. This position involves working closely with frontend, backend, and infrastructure teams to deliver the Tinker console, developer tools, and essential features. What you will do Develop and enhance Tinker’s APIs and backend services using Python and Rust, focusing on areas like job submission, orchestration, billing, and usage tracking. Design and launch user interfaces, including the Tinker console and upcoming developer tools, using React and TypeScript. Refine the developer experience by improving SDK usability, error messages, API design, and onboarding processes. Work to increase system reliability, observability, and security in production, and participate in on-call rotations. Create internal tools that help research and infrastructure teams work more efficiently. Location This role is based in San Francisco, California.

Thinking Machines Lab brings together scientists, engineers, and innovators who have shaped well-known AI products like ChatGPT and Character.ai, as well as open-weight models such as Mistral. The team also contributes to open-source projects including PyTorch, OpenAI Gym, Fairseq, and Segment Anything. The company’s mission centers on advancing collaborative general intelligence, aiming to make AI accessible and adaptable to individual needs. Tinker, the company’s fine-tuning API, enables researchers and developers to customize advanced AI models using their own data and algorithms. Thinking Machines manages the infrastructure, giving users the flexibility to train open-weight models while focusing on their unique requirements. As Tinker expands, the platform continues to evolve alongside its growing community. Role overview The Site Reliability Engineer will focus on improving the reliability and resilience of the Tinker platform. This role involves close collaboration with platform engineers and research teams to strengthen every layer of the system, from infrastructure to user-facing services. What you will do Define and take ownership of end-to-end reliability, including CI/CD workflows, production observability, and incident response processes. Set Service Level Objectives for distributed training systems, balancing reliability, scheduling latency, and development speed. Design and implement monitoring and observability across the training pipeline. Manage incident response for Tinker, ensuring prompt recovery, thorough incident analysis, and systematic improvements to prevent recurrence. Enhance multi-tenant isolation and resource scheduling to support LoRA-based workload co-scheduling, maintaining both reliability and data separation. Collaborate with security teams to identify and address production vulnerabilities. This position is based in San Francisco.

Thinking Machines Lab brings together scientists, engineers, and innovators with a track record in developing widely used AI products and open-source projects. The team has contributed to tools like ChatGPT, Character.ai, Mistral, PyTorch, OpenAI Gym, Fairseq, and Segment Anything. The company’s mission centers on advancing collaborative general intelligence to help people achieve more with AI tailored to their needs. Tinker, the company’s fine-tuning API, enables researchers and developers to adapt advanced AI models to their own data and algorithms. By handling the infrastructure, Tinker allows users to focus on customization, opening up capabilities that were once limited to a few specialized labs. As Tinker’s customer base and feature set grow, the team is focused on building a scalable platform and supporting an expanding community. Role overview The GTM Strategy & Operations Lead will build and refine the commercial structure for Tinker. This person will design strategies and processes that turn organic product adoption into a consistent, scalable revenue stream. The role involves shaping how Tinker’s fine-tuning capabilities are packaged, priced, launched, and sold across different customer segments. Collaboration with product, engineering, and research teams is central to the work. Tinker is designed for technically sophisticated users. The GTM lead must be comfortable discussing training infrastructure and understand how developers evaluate and adopt new tools. What you will do Develop and execute commercialization strategies for Tinker, including pricing, packaging, and launch plans based on market and competitor analysis. Create go-to-market approaches tailored to different types of customers. Manage partnerships to expand Tinker’s reach and open new channels for demand. Design and oversee customer pilots, onboarding, and expansion playbooks to move accounts from testing to production use. Produce commercial playbooks to help customer-facing engineers and FDEs position and sell Tinker effectively. Set and track success metrics for launches and GTM projects, running experiments to test assumptions about pricing and product packaging.

Thinking Machines Lab brings together scientists, engineers, and innovators who have contributed to well-known AI products such as ChatGPT, Character.ai, and open-source frameworks like PyTorch, OpenAI Gym, Fairseq, and Segment Anything. The team's mission centers on advancing collaborative general intelligence, aiming to make AI accessible for people to address their own needs and ambitions. The Tinker platform offers a fine-tuning API that lets researchers and developers tailor advanced AI models to their specific requirements. Tinker provides the infrastructure, while users maintain flexibility to train open-weight models with their own data and algorithms. As Tinker grows its features and user base, the team is expanding to support the platform's evolution. Role overview This Full Stack Software Engineer role focuses on designing, building, and maintaining the products and services that Tinker users rely on. The work covers frontend, backend, and infrastructure, with an emphasis on the Tinker console, developer tools, and meeting the changing needs of the Tinker community. What you will do Develop and improve Tinker’s APIs and backend services using Python and Rust, including systems for job submission, orchestration, billing, and usage tracking. Build user-facing interfaces such as the Tinker console and future developer tools with React and TypeScript. Enhance the developer experience by refining SDK usability, error messages, API design, and onboarding workflows. Increase system reliability, observability, and security in Tinker’s production environment, and participate in on-call rotations. Create internal tools to support the research and infrastructure teams working on Tinker. This position is based in San Francisco.

Product ResearcherAs a Product Researcher at Firecrawl, you will be pivotal in shaping our future product development. Our engineering team is currently focused on delivering robust infrastructure, while our sales and support teams effectively manage incoming queries. However, we are seeking a dedicated individual to drive our larger product initiatives—those that will elevate our tool into a leading platform. Currently, we possess a cutting-edge commercial research paper search endpoint with limited use. Our features, including answers, reranking, and monitoring, require the dedicated focus of a researcher like you. You will ensure that we prioritize effectively based on genuine customer needs.This role transcends traditional product management; you will immerse yourself in the field, engaging in frequent conversations with customers—conducting 15-25 interactions weekly. These will include discovery calls, user interviews, onboarding sessions, and discussions surrounding churn. Your mission is to uncover insights that reveal our customers' challenges and aspirations, effectively bridging the gap between our developers' efforts and our product roadmap.

At Thinking Machines Lab, our mission is to enhance humanity's potential through the advancement of collaborative general intelligence. We envision a future where everyone can access the knowledge and tools necessary to leverage AI for their individual needs and objectives.As a team of scientists, engineers, and innovators, we have developed some of the most widely utilized AI products, such as ChatGPT and Character.ai, along with open-weight models like Mistral, and popular open-source projects including PyTorch, OpenAI Gym, Fairseq, and Segment Anything.Role OverviewAt Thinking Machines, we adopt a multimodal-first approach, where multimodality is integral to our scientific goals and infrastructure. We are seeking skilled researchers to push the boundaries of audio capabilities. In this role, you will delve into how audio models can facilitate more natural and efficient communication and collaboration by preserving information and accurately capturing user intent.This position requires strong collaboration across pre-training, post-training, and product development with top-tier researchers, infrastructure engineers, and designers. Here, you will have the chance to influence the foundational capabilities of AI systems that will be utilized by millions globally.This role marries fundamental research with practical engineering, as we do not separate these functions internally. You will be expected to write high-performance code as well as engage with technical reports. It is an ideal position for someone who enjoys both extensive theoretical research and hands-on experimentation while laying the groundwork for how AI learns.Note: This is an evergreen role that remains open continuously to gauge interest in this research area. We receive numerous applications, and there may not always be an immediate match for your skills and experience. Nevertheless, we encourage you to apply, as we routinely review applications and reach out as new opportunities arise. You are welcome to reapply if you gain additional experience, but please avoid applying more frequently than every six months. Occasionally, we also post specific roles for distinct project or team needs, in which case you are welcome to apply directly alongside an evergreen submission.

At Thinking Machines Lab, we are dedicated to empowering humanity through the advancement of collaborative general intelligence. Our vision is to create a future where everyone has access to the knowledge and tools necessary to harness AI for their distinct needs and aspirations.Our team comprises scientists, engineers, and innovators who have developed some of the most utilized AI products globally, including ChatGPT and Character.ai, as well as leading open-weight models like Mistral and popular open-source initiatives such as PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the RoleAs the Lead for Data Partnerships at Thinking Machines Lab, you will oversee the complete data procurement pipeline for frontier model training. This includes understanding the data requirements of our research teams, sourcing and finalizing agreements with providers, and managing the quality and delivery of data. You will serve as the bridge connecting our research, legal teams, and external vendors, ensuring timely access to the right data for our teams.This role is perfect for someone with a technical inclination who is eager to delve into the intricacies of data to support an ambitious research agenda. You must be adept at switching contexts between planning the data needed for training runs and negotiating pricing with vendors. Over time, you will establish scalable and repeatable processes to ensure our data operations align with the pace of our research efforts.What You Will DoLead and coordinate end-to-end data procurement initiatives, ensuring complex sourcing activities are conducted with efficiency, transparency, and scientific rigor.Collaborate closely with research teams to proactively identify data needs across pre-training, post-training, and evaluation workstreams, anticipating requirements rather than merely reacting to requests.Source, assess, and onboard data providers, developing and maintaining a pipeline of potential vendors across various domains.Negotiate pricing, licensing terms, and contract structures with data providers, collaborating with legal teams to finalize agreements that align with our research objectives.Evaluate incoming data alongside researchers, determining quality and coverage for intended training goals.Monitor and manage ongoing data deliveries, tracking schedules, addressing issues, and verifying that received data aligns with agreements.Create repeatable, scalable processes surrounding the entire data procurement lifecycle, enhancing the speed and systematic nature of data sourcing over time.Translate technical data requirements into actionable plans with clear milestones, ensuring team alignment across projects.

At Thinking Machines Lab, our mission is to empower humanity by advancing collaborative general intelligence. We envision a future where everyone has access to the knowledge and tools necessary to make AI work for their individual needs and goals. Our team comprises scientists, engineers, and innovators who have developed some of the most widely adopted AI products, including ChatGPT and Character.ai, alongside open-weight models like Mistral, as well as popular open-source initiatives such as PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the RoleWe are seeking a highly skilled infrastructure research engineer to architect and develop core systems that facilitate efficient large-scale model training, with a strong emphasis on numerics. You will enhance the numerical foundations of our distributed training stack, focusing on precision formats, kernel optimizations, and communication frameworks to ensure that training trillion-parameter models is stable, scalable, and fast.This position is perfect for an individual who excels at the intersection of research and systems engineering—a creator who comprehends both the mathematics of optimization and the practicalities of distributed computing.Note: This is an "evergreen role" that remains open for ongoing expressions of interest. While we receive numerous applications and there may not always be an immediate opening that perfectly matches your skills and experience, we encourage you to apply. We continuously review applications and will contact applicants as new opportunities arise. You are welcome to reapply if you gain additional experience, but please refrain from applying more than once every six months. You may also notice postings for specific roles related to particular projects or teams; in those instances, you are welcome to apply for those positions in addition to the evergreen role.What You’ll DoDesign and optimize distributed training infrastructure for large-scale LLMs, ensuring performance, stability, and reproducibility in multi-GPU and multi-node environments.Implement and assess low-precision numerics (e.g., BF16, MXFP8, NVFP4) to enhance efficiency while maintaining model quality.Develop kernels and communication primitives that leverage hardware-level support for mixed and low-precision arithmetic.Collaborate with research teams to co-design model architectures and training methodologies that align with new numeric formats and stability requirements.Prototype and benchmark scaling strategies, including data, tensor, and pipeline parallelism that integrate precision-adaptive computation and quantized communication.Contribute to the design of our internal orchestration and monitoring frameworks.

Join the Center for AI Safety (CAIS), a pioneering research and advocacy organization dedicated to addressing the societal-scale risks posed by artificial intelligence. We tackle the most pressing challenges in AI through rigorous technical research, innovative field-building initiatives, and proactive policy engagement, in collaboration with our sister organization, the Center for AI Safety Action Fund.As a Research Scientist, you will spearhead and conduct transformative research aimed at enhancing the safety and dependability of cutting-edge AI systems. Your responsibilities will include designing and executing experiments on large language models, developing the necessary tools for training and evaluating models at scale, and converting your findings into publishable research. You will work closely with CAIS researchers and external partners from academia and industry, utilizing our compute cluster for large-scale model training and evaluation. Your research will focus on critical areas such as AI honesty, robustness, transparency, and the detection of trojan/backdoor behaviors, all aimed at mitigating real-world risks associated with advanced AI technologies.

About MercorMercor sits at the forefront of labor markets and artificial intelligence research, collaborating with premier AI laboratories and enterprises to harness the human intelligence crucial for AI evolution.Our expansive talent network empowers the training of cutting-edge AI models, akin to how educators impart knowledge to students—sharing insights, experiences, and contexts that transcend mere code. Currently, our network comprises over 30,000 experts, generating collective earnings exceeding $2 million daily.At Mercor, we are pioneering a unique category of work where expertise fuels AI progress. Realizing this vision necessitates a bold, fast-paced, and deeply dedicated team. You will collaborate with researchers, operators, and AI firms that are at the vanguard of transforming systems that redefine society.As a profitable Series C company, Mercor is valued at $10 billion and maintains an in-office presence five days a week at our new headquarters in San Francisco.About the RoleIn your capacity as a Research Engineer at Mercor, you will operate at the intersection of engineering and applied AI research. You will play a pivotal role in post-training and Reinforcement Learning from Human Feedback (RLVR), synthetic data generation, and large-scale evaluation workflows essential for advancing frontier language models.Your contributions will help train large language models to adeptly utilize tools, exhibit agentic behavior, and engage in real-world reasoning within production environments. You will be instrumental in shaping rewards, conducting post-training experiments, and constructing scalable systems to enhance model performance. Your responsibilities will also include designing and evaluating datasets, creating scalable data augmentation pipelines, and developing rubrics and evaluators that expand the learning potential of LLMs.

At Thinking Machines Lab, our mission is to empower humanity by advancing collaborative general intelligence. We envision a future where everyone has access to the knowledge and tools necessary to harness AI for their unique needs and goals.Our team comprises scientists, engineers, and builders who have developed some of the most widely utilized AI products, such as ChatGPT and Character.ai, alongside open-weight models like Mistral, and popular open-source initiatives like PyTorch, OpenAI Gym, Fairseq, and Segment Anything.About the PositionWe are seeking an Infrastructure Research Engineer to design and construct the foundational systems that facilitate the scalable and efficient training of large models for both deployment and research purposes. Your primary objective will be to streamline experimentation and training at Thinking Machines, enabling our research teams to concentrate on scientific advancements rather than system limitations.This role is a perfect match for an individual who possesses a strong blend of deep systems expertise and a keen interest in machine learning at scale. You will take full ownership of the training stack, ensuring that every GPU cycle contributes to scientific progress.Note: This is an evergreen role that we keep open continuously to express interest. We receive numerous applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. However, we encourage you to apply. We regularly review applications and reach out to candidates as new opportunities arise. Feel free to reapply as you gain more experience, but please avoid applying more than once every six months. We may also post specific roles for individual projects or team needs, in which case you are welcome to apply directly alongside this evergreen role.Key ResponsibilitiesDesign, implement, and optimize distributed training systems that scale across thousands of GPUs and nodes for extensive training workloads.Develop high-performance optimizations to maximize throughput and efficiency.Create reusable frameworks and libraries that enhance training reproducibility, reliability, and scalability for new model architectures.Establish standards for reliability, maintainability, and security, ensuring systems remain robust under rapid iterations.Collaborate with researchers and engineers to construct scalable infrastructure.Publish and disseminate findings through internal documentation, open-source libraries, or technical reports that contribute to the advancement of scalable AI infrastructure.

About the RoleJoin Thorin as an AI Researcher and play a pivotal role in shaping the core research initiatives that drive our AI innovations. In this position, you will operate at the crossroads of machine learning research, practical model development, and product application, enhancing our understanding and automation of enterprise workflows.This position merges theoretical research with hands-on implementation, transitioning ideas from conceptual stages through experimentation into functional components that enrich Thorin’s offerings.Your ResponsibilitiesResearch & InnovationConduct innovative machine learning research aligned with real-world product demands.Investigate new model architectures, training methods, and evaluation techniques specifically designed for understanding and automating organizational workflows.Model Development & EvaluationCreate, implement, and assess ML/AI methodologies that enhance model efficacy for essential tasks.Collaborate closely with cross-functional teams to integrate research outcomes into tangible products that meet user needs.

At magic.dev, we are committed to advancing humanity by developing safe artificial general intelligence (AGI) that tackles the world's most pressing challenges. Our unique approach focuses on automating research and code generation to enhance model performance and alignment more effectively than traditional methods. By leveraging cutting-edge pre-training, domain-specific reinforcement learning, ultra-long context processing, and efficient inference-time computation, we aim to redefine the capabilities of AGI.Role OverviewAs a Research Engineer, you will play a pivotal role in training, evaluating, and deploying large-scale AI models alongside innovative inference-time computing methods. You will contribute to the creation of extensive internet-scale datasets and support the prototyping of groundbreaking research and product initiatives.Key ResponsibilitiesEnhance inference throughput for cutting-edge model architecturesDevelop and refine frameworks that underpin our research and production processesTrain trillion-parameter models using large GPU clustersCurate post-training datasets to bolster specific capabilitiesConstruct internet-scale data pipelines and web crawlersDesign, prototype, and optimize innovative model architecturesContribute to cutting-edge research in long-context, inference-time computation, reinforcement learning, and additional domainsQualificationsProven software engineering expertiseIn-depth understanding of deep learning literatureExperience with both pre-training and post-training of large language models (LLMs)Strong capability to generate and assess research ideasFamiliarity with large distributed systemsProficient in managing substantial ETL workloadsCompensation and BenefitsAnnual salary ranging from $225,000 to $550,000 based on experienceEquity is a significant component of total compensation401(k) plan with a 6% salary matchComprehensive health, dental, and vision insurance for you and your dependentsUnlimited paid time offVisa sponsorship and relocation assistance availableBe part of a small, dynamic, and focused team