Research Intern in San Francisco

About the Team OpenAI's Human Data Team creates custom data solutions driving groundbreaking research. Our work enhances and evaluates our flagship models and products like ChatGPT, GPT-5, and Sora, and contributes to safety initiatives through collaboration with our Preparedness and Safety Systems teams. About the Role As a Research Program Manager (RPM) in the Human Data team you will partner with research and engineering to design and implement pragmatic solutions for collecting high-quality data. You will be a key interface between our research roadmap, external vendors, AI trainers, and the Human Data engineering team. This role is based in our San Francisco HQ. In this role, you will: Collaborate with Research: Partner with researchers to scope data collection needs, define success metrics, and establish quality measurement frameworks. Design & Execute Data Collection Campaigns: Translate research needs into actionable plans and accelerate execution by leveraging existing tooling and iterating to reach the desired outcome. In many cases, you will need to implement scrappy new solutions while partnering with engineering to design robust/scalable solutions. Unblock Yourself: You must be deeply uncomfortable with the idea of sitting around waiting for external dependencies, and have the technical acumen and drive to figure out how to achieve at least partial success in the interim. Optimize Systems & Processes: Build and optimize dashboards to track campaign performance, leveraging SQL and Python for data analysis and actionable insights. Drive Technical Roadmaps: Collaborate with engineers to enhance data platforms, resolve blockers, and ensure security best practices such as access management. Scale Your Impact : Advise and empower program managers and vendors to drive day-to-day execution so that you can focus on addressing high priority opportunities. You might thrive in this role if you: Are proficient in SQL and Python for data analysis, including q

About the team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a researcher working on Frontier Evals & Environments, you will help build north star model environments to drive progress towards safe AGI/ASI. Your work will directly guide the research programs of the most ambitious training runs happening at OpenAI. Some prior open-sourced evaluations built by researchers in this role include GDPval , SWE-bench Verified , MLE-bench , PaperBench , and SWE-Lancer . If you are interested in feeling firsthand the fast progress of our models, and steering them towards good outcomes, this is the role for you. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measure whether it worked, and ship improvements into products used by real people. This is a high-agency role for people who want their work to land directly in frontier models. In this role, you'll: Create ambitious RL environments to push our models to their limits, and measure frontier

About the Team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a member of Agent Post-Training, Connectors, you will teach models how to interface with the top professional software using code. You will help train agents to use code, APIs, tools, and structured integrations to operate across applications like Slack, Google Workspace, GitHub, Notion, Linear, Salesforce, and other core systems of work. You will help enable models to take useful actions across a user’s digital context: finding information, updating systems, coordinating work, generating artifacts, and completing multi-step workflows through the tools teams already use. You will train models to be supercharged by the world’s most important productivity and enterprise software, turning connected tools into a powerful action surface for our agents. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measure whether it worked, and ship improvements into products used by real people.

About the Team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a member of Agent Post-Training, you will improve the capabilities, reliability, and product fit of OpenAI's agentic models. You might own a research direction, build the infrastructure that makes large training runs faster and more trustworthy, create evals that reveal where models fail, or drive a capability from an idea through experimentation, integration, and launch. This role is intentionally broad. The strongest candidates are not defined by one method or subfield; they are people who can take an ambiguous capability problem and make progress across research, engineering, data, evals, and product. You should be excited to work on models that act in the world: writing and debugging code, using tools, calling functions, operating computers, collaborating with other agents, and completing valuable work on behalf of users. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, meas

About the Team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a researcher working on Frontier Evals & Environments, you will help build north star model environments to drive progress towards safe AGI/ASI. Your work will directly guide the research programs of the most ambitious training runs happening at OpenAI. Some prior open-sourced evaluations built by researchers in this role include GDPval , SWE-bench Verified , MLE-bench , PaperBench , and SWE-Lancer . If you are interested in feeling firsthand the fast progress of our models, and steering them towards good outcomes, this is the role for you. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measure whether it worked, and ship improvements into products used by real people. This is a high-agency role for people who want their work to land directly in frontier models. In this role, you might Create ambitious RL environments to push our models to their limits, and measure frontie

About the Team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role We believe that the final enabler for AGI is spending compute on context. As a Context Researcher on Agent Post-Training, you will scale compute spent on context. You will get to work in our frontier training stack on enabling the next paradigm of model training with a clear product interface for iterative deployment (Codex Chronicle). You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measure whether it worked, and ship improvements into products used by real people. This is a high-agency role for people who want their work to land directly in frontier models. In this role, you might Design and run experiments that improve scaling of compute on context. Own end-to-end improvements to the post-training stack, including RL, data pipelines, graders, reward signals, evals, diagnostics, and model-behavior analysis. Build evals and environments that expose the next set of model failures,

About the Team The Agent Post-Training team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a member of Agent Post-Training, Artifacts, you will train frontier models to create polished, useful work products: documents, spreadsheets, slide decks, dashboards, reports, analyses, and other interactive or editable artifacts. You will help teach our models to move from a vague user goal to a finished artifact with strong structure, visual taste, domain judgment, correctness, and low latency. This work will require owning improvements across our post-training stack, including RL, data pipelines, graders, reward signals, evals, and behavioral analysis. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measure whether it worked, and ship improvements into products used by real people. This is a high-agency role for people who want their work to land directly in frontier models. In this role, you will: Design and run experiments that improve agentic model behavior for complex so

About the Team The Codex Research team creates the frontier agents OpenAI ships to the world. We are training the models behind our agents in Codex, ChatGPT, the API, and other frontier products: persistent, proactive intelligence that can operate computers, collaborate with people and other agents, and expand what people and organizations can imagine, attempt, and achieve. We define what the next generation of agents should be able to do, build the training signal that teaches those abilities, and run the experiments that make them real. Our work spans coding, tool use, computer use, multi-agent coordination, long-horizon execution, factuality, instruction following, calibrated reasoning, and taste. Our team is where new model capabilities get made. We build the data, environments, graders, training methods, and feedback loops that shape what OpenAI's next agents can do, then carry those capabilities through major training runs and into the products people use. About the Role As a member of the Codex Research team, you will improve the capabilities, reliability, and product fit of OpenAI's agentic models. You might own a research direction, build the infrastructure that makes large training runs faster and more trustworthy, create evals that reveal where models fail, or drive a capability from an idea through experimentation, integration, and launch. This role is intentionally broad. The strongest candidates are not defined by one method or subfield; they are people who can take an ambiguous capability problem and make progress across research, engineering, data, evals, and product. You should be excited to work on models that act in the world: writing and debugging code, using tools, calling functions, operating computers, collaborating with other agents, and completing valuable work on behalf of users. You will work with researchers, engineers, product teams, infrastructure teams, and safety/alignment partners to decide what should go into major model runs, measu

About the Team OpenAI’s mission is to ensure that general-purpose artificial intelligence benefits all of humanity. Our Communications team’s ethos is to support OpenAI’s mission and goals by clearly and authentically explaining our technology, values, and approach to safely building powerful AI. About the Role OpenAI is seeking an experienced communications professional to join our Platform & Research Communications team. This role will work closely with the Research Communications Lead and partner deeply with safety researchers, alignment researchers, and cross-functional teams to shape how OpenAI’s safety research is understood by researchers, journalists, policymakers, and the broader public. This position is responsible for developing and executing external communications strategies around OpenAI’s safety research—from alignment and evaluations to broader work that helps advance the safe development and deployment of increasingly capable AI systems. The ideal candidate brings strong science or technical fluency, excellent storytelling instincts, and experience helping researchers communicate complex work with clarity, accuracy, and nuance. You will partner closely with research leadership, individual researchers, policy, product, safety, legal, and cross-functional communications teams. This role requires both strategic judgment and hands-on execution in a fast-moving environment where research, public understanding, and high-stakes safety narratives intersect. This role is based in San Francisco, CA and follows a hybrid schedule (three days per week in office). Relocation assistance is available. In this role, you will: Shape Safety Research Narratives Develop clear, credible external narratives around OpenAI’s safety research, including alignment, evaluations, preparedness, interpretability, and other areas connected to the safe development of frontier AI. Translate complex technical work into accessible stories without oversimplifying, overstating impact

About the Team The Workload team is responsible for designing and running OpenAI’s LLM training and inference infrastructure that powers frontier models at massive scale. Our systems unify how researchers train and serve models, abstracting away the complexity of performance, parallelism, and execution across vast GPU/accelerator fleets. By providing this foundation, the Workload team ensures that researchers can focus on advancing model capabilities while we handle the scale, efficiency, and reliability required to bring those models to life. About the Role We are looking for an engineer to design and implement the dataset infrastructure that powers OpenAI’s next-generation training stack. You will be responsible for building standardized dataset interfaces, scaling pipelines across thousands of GPUs, and proactively testing performance bottlenecks. In this role, you will collaborate closely with the multimodal researchers, and other infra groups to ensure datasets are unified, efficient, and easy to consume. In this role, you will: Design and maintain standardized dataset APIs, including for multimodal (MM) data that cannot fit in memory. Build proactive testing and scale validation pipelines for dataset loading at GPU scale. Collaborate with teammates to integrate datasets seamlessly into training and inference pipelines, ensuring smooth adoption and a great user experience. Document and maintain dataset interfaces so they are discoverable, consistent, and easy for other teams to adopt. Establish safeguards and validation systems to ensure datasets remain reproducible and unchanged once standardized. Debug and resolve performance bottlenecks in distributed dataset loading (e.g., straggler systems slowing global training). Provide visualization and inspection tools to surface errors, bugs, or bottlenecks in datasets. You might thrive in this role if you: Have strong engineering fundamentals with experience in distributed systems, data pipelines, or infrastructure.

About the Team Our Robotics team is focused on unlocking general-purpose robotics and pushing towards AGI-level intelligence in dynamic, real-world settings. Working across the entire model stack, we integrate cutting-edge hardware and software to explore a broad range of robotic form factors. We strive to seamlessly blend high-level AI capabilities with the constraints of physical systems to improve peoples’ lives. About the Role We’re looking for a Robotics Control Systems Engineer to take on a foundational role within our robotics team. You’ll help architect, implement, tune, and verify the control infrastructure that enables intelligent, reliable, and responsive robot behavior. This is a deeply hands-on role focused on real-time systems, actuation, dynamics, low-level hardware interaction, and whole-robot performance. You’ll spend significant time working directly with robots onsite: debugging behavior, tuning subsystems, running experiments, and providing feedback across mechanical, electrical, and software teams. This role is based in San Francisco, CA, and requires in-person 5 days a week. In this role, you will: Design and implement real-time control algorithms for robotic systems, including motion control, feedback loops, state estimation, actuator control, and subsystem tuning. Define the control architecture from low-level actuators and hardware interfaces through whole-robot behavior and policy. Identify and characterize actuator, hardware, and software parameters through rigorous experimentation, testing, commissioning, and verification. Work with machine learning engineers to implement reinforcement learning models. Collaborate across mechanical, electrical, and software teams to integrate control logic with sensing and actuation hardware. Help inform the mechanical and electrical design to maximize capability and flexibility. Create the control system architecture; determine the correct level of abstraction from actuators all the way up to whole-robot

About the Team Our Robotics team is focused on unlocking general-purpose robotics and pushing towards AGI-level intelligence in dynamic, real-world settings. Working across the entire model stack, we integrate cutting-edge hardware and software to explore a broad range of robotic form factors. We strive to seamlessly blend high-level AI capabilities with the constraints of physical systems to improve peoples’ lives. About the Role We are seeking an Electrical Engineer to build and own the electrical backbone of our robotic actuator dynamometer and test infrastructure. You will design, integrate, and operate the load motor drives, power distribution, instrumentation wiring, DAQ interfaces, and safety systems that make high-performance robotic actuator testing repeatable, safe, and scalable. This role spans hands-on lab execution and system architecture: selecting and commissioning power electronics, designing robust test-cell electrical systems, bringing up sensors and DAQ, and partnering with mechanical and software engineers to turn robotic actuator hardware into trustworthy data. In this role, you will Own the electrical architecture of dynamometer and actuator test cells, from mains distribution and protection through load motor drives, braking, and auxiliary power. Specify, integrate, commission, and tune motor drives and load machines for robotic actuator torque, speed, efficiency, thermal, and durability testing. Design power distribution, grounding, shielding, cable routing, and connectorization for high-current, high-voltage, and low-level measurement systems. Integrate torque, position, speed, temperature, voltage, current, vibration, and other instrumentation from robotic actuators into DAQ and control systems. Develop electrical schematics, wiring diagrams, panel layouts, harness documentation, and test-cell interface definitions. Build, debug, and maintain test-cell electrical hardware, rapidly diagnosing noise, EMI, grounding, drive, sensor, and power-q

About the Team The Alignment team at OpenAI is dedicated to ensuring that our AI systems are safe, trustworthy, and consistently aligned with human values, even as they scale in complexity and capability. Our work is at the cutting edge of AI research, focusing on developing methodologies that enable AI to robustly follow human intent across a wide range of scenarios, including those that are adversarial or high-stakes. We concentrate on the most pressing challenges, ensuring our work addresses areas where AI could have the most significant consequences. By focusing on risks that we can quantify and where our efforts can make a tangible difference, we aim to ensure that our models are ready for the complex, real-world environments in which they will be deployed. The two pillars of our approach are: (1) harnessing improved capabilities into alignment, making sure that our alignment techniques improve, rather than break, as capabilities grow, and (2) centering humans by developing mechanisms and interfaces that enable humans to both express their intent and to effectively supervise and control AIs, even in highly complex situations. About the Role As a Research Engineer / Research Scientist on the Alignment team, you will be at the forefront of ensuring that our AI systems consistently follow human intent, even in complex and unpredictable scenarios. Your role will involve designing and implementing scalable solutions that ensure the alignment of AI as their capabilities grow and that integrate human oversight into AI decision-making. This role is especially well suited for someone who can move from an ambiguous model-behavior question to a concrete experimental setup: formulate the hypothesis, build the evaluation or intervention, run the experiment, analyze the result, and decide what the evidence supports. This role may be based in San Francisco or London, subject to team needs and location approval. In this role, you will: We are seeking research engineers and res

About the Team OpenAI’s mission is to build safe artificial general intelligence (AGI) that benefits all of humanity. Achieving this requires bringing the world’s most exceptional talent under one roof to push the boundaries of what’s possible. Our Research Recruiting team plays a critical role in this effort. We are an embedded part of the research organization—working side by side with our research staff to deeply understand evolving priorities, build trust, and strategically shape the future of OpenAI’s talent. About the Role We are looking for a highly strategic recruiter to work closely with the Head of Research Recruiting on a small set of unusually important, high-touch searches and candidate relationships. This role will focus on exceptional talent who does not move through a standard recruiting process: highly visible researchers, technical leaders, operators, and other special-interest candidates where timing, discretion, market intelligence, and tailored engagement matter as much as process execution. This is not a conventional req-based recruiting role. You will help identify where the market is moving, develop intelligence on top talent and competitor activity, translate that intelligence into action, and orchestrate bespoke recruiting strategies for candidates who require a more nuanced path into OpenAI. You should be able to translate these signals and states into clear, actionable advice for leaders and then make it happen. In this role, you will: Partner directly with the Head of Research Recruiting and research leadership team to define priority talent targets and shape bespoke engagement strategies. Build and maintain deep market intelligence across frontier AI, research, engineering, and adjacent talent ecosystems, including competitor movement, candidate motivations, and relationship context. Proactively identify, map, and cultivate exceptional high-profile talent before there is a formal or standard hiring process attached. Translate weak signa

About the Team We are building general-purpose robotics. In the short term, we are focused on robots to support skilled workers to build our future infrastructure. In the long term, we imagine everyone having a personal robot doing anything they need. Progress is rapid, and based on a foundation of co-design between robotics hardware and ML research. About the Role As a Firmware Engineer, you will define and drive the architecture of embedded systems for next-generation hardware products. You will own foundational firmware decisions across real-time execution, device bring-up, hardware interfaces, fault handling, safety mechanisms, and production readiness. We’re looking for someone with deep experience building safety-critical or high-consequence systems, where failures can have meaningful consequences. You should be comfortable reasoning about risk, designing for diagnosability and graceful degradation, and creating engineering practices that raise the reliability bar for the entire team. You should also be unusually good at moving fast. Sometimes the right answer is a carefully reviewed architecture that will endure for years; sometimes it is getting a rough-but-useful prototype working by the end of the afternoon so the team can learn something concrete tomorrow. We value engineers who know the difference, make that call well, and can operate credibly in both modes. You will be both a technical leader and a hands-on builder: setting direction, reviewing critical designs, unblocking the hardest problems, and writing production firmware when it matters most. Our embedded stack uses a lot of Rust. Extensive experience in the language is a big help! This role is based in San Francisco, CA. This role will be expected to be in office 4 days per week and offer relocation assistance to new employees. In this role, you will: Rapidly bring up new hardware and set execution pace for the team. Lead firmware architecture for embedded systems spanning boot, RTOS/runtime behav