Software Engineer, Model Capabilities

About the Team OpenAI’s Hardware organization develops AI-native silicon and system-level solutions for the unique demands of advanced AI workloads. Building on efforts like Jalapeño, the team is developing future generations of AI-native silicon and tightly integrated systems to power the next generation of frontier models. By co-designing chips, systems, tools, and methodologies, the team helps deliver faster, more efficient, and production-ready hardware for OpenAI’s supercomputing platform. About the Role You will build the model runtime within the inference engine that executes complex, frontier models at scale on OpenAI’s custom silicon. The runtime will sit between models running on the hardware and the upper layers of the cluster serving software stack, translating demanding inference workloads into efficient execution while optimizing for throughput, latency, utilization, and reliability. You will work across model architecture, distributed systems, compilers, kernels, and silicon to design a production-grade runtime comparable in ambition to systems such as vLLM and SGLang, but customized and optimized for OpenAI’s AI accelerator. Your work will shape how new model capabilities map onto the platform and how quickly custom silicon can deliver meaningful performance in production. In this role, you will: Design and implement the LLM inference runtime for frontier models running on custom silicon. Build scheduling, continuous batching, memory management, KV-cache management, and execution orchestration for high-performance inference. Develop distributed execution strategies across chips, hosts, and racks, including model partitioning, communication, and synchronization. Optimize end-to-end latency, throughput, memory efficiency, and hardware utilization across diverse model architectures and serving workloads. Partner with kernel, compiler, architecture, and silicon teams to co-design interfaces and remove performance bottlenecks across the stack. Enable new

About the Team API Multimodal builds the developer-facing products and infrastructure that bring OpenAI’s image, audio, and real-time model capabilities into the world. We are responsible for high-scale APIs for image generation, speech transcription, speech generation, and low-latency voice interactions. We partner closely with Research and Inference to bring frontier model capabilities to developers and use customer feedback to improve our models. About the Role As a software engineer on API Multimodal, you will build and operate the products and distributed systems behind OpenAI’s image, audio, and real-time APIs. You will work across model integration, API design, and production infrastructure to turn new research capabilities into reliable developer experiences. This hands-on role combines backend and systems depth with product judgment: you will own projects end to end, partner with Research, Inference, and Safety, and help make multimodal AI useful at scale. Model training experience is not required. In this role, you will: Design, build, and ship developer-facing APIs and backend services that serve frontier models. Architect low-latency streaming, request, session, and model integration systems that make complex multimodal interactions reliable and intuitive at scale. Work directly with Research to bring new model capabilities into production, shape the systems around them, and incorporate feedback from real-world developers and customers. Own the availability, latency, scalability, and cost efficiency of the services you build. Own projects from technical design and implementation through launch and ongoing iteration, while raising the team’s engineering standards. Your background might look something like: 7+ years of professional experience, excluding internships, in backend, infrastructure, platform, or product engineering roles. A track record of designing, building, and operating production backend services, developer-facing APIs, or distributed syste

About the team The Computer Use and New Interfaces team is focused on discovering and building the next generation of AI-native interfaces. We believe that the value of AI is increasingly constrained not by model capabilities, but by the ways people interact with those capabilities. Our mission is to create new interaction paradigms that unlock the full potential of AI and integrate it more deeply into people's lives and work. Our team does both near-term product development and longer-term product incubation that can influence experiences across ChatGPT, Codex, future OpenAI products, and emerging device platforms. We work in a highly collaborative, design-driven environment where engineering, product, and design operate as one team. We value rapid experimentation, prototyping, and iteration, creating the shortest possible path between an idea, a working system, and a product decision. About the role We're looking for exceptional engineers who are excited to invent entirely new ways for people to interact with AI. You'll work at the intersection of engineering, product, and design to explore, prototype, and build novel interface concepts that push beyond traditional software paradigms. This role requires comfort with ambiguity, strong product instincts, and a willingness to move fluidly between experimentation and production systems. You'll help shape both the capabilities and the user experiences that define how people engage with AI. In this role you will: Design, prototype, and build novel AI-native interfaces and interaction models. Develop foundational technologies and frameworks that enable generative UI and computer use experiences. Collaborate closely with designers, product thinkers, and engineers to rapidly explore and validate new concepts. Build end-to-end prototypes and production systems that can influence future OpenAI products. Contribute to platform technologies that can be adopted across multiple product surfaces. Help establish technical directio

About the Team The Emerging Products team is a lean, high-output product lab group that builds products at the forefront of model capabilities. We collaborate across all teams within the company, from research and infrastructure to consumer products. The team is responsible for identifying new product opportunities, building them quickly, dogfooding them internally, and then launching the successful products to users. We use data, user research, and analytics to inform our ideas, and make decisions on what experiments are worth iterating, stopping, or scaling. About the Role We’re looking for a senior, product-minded software engineer to own ambiguous 0-to-1 work from idea through prototype, validation, and handoff. This is a full-stack role with a strong frontend and product emphasis: you will build the interfaces and supporting backend systems needed to test new experiences quickly, while making sound architectural choices that enable successful concepts to scale. This role is based in our Mission Bay office in San Francisco. In this role, you will: Build and ship high-quality, product experiments across the full stack. Turn ambiguous user needs and emerging technical capabilities into testable product concepts, using research and metrics to guide iteration. Own technical direction for 0-to-1 projects, balancing speed, reliability, and a clear path from prototype to scalable product. Partner closely with design, product, research, and engineering teams to dogfood, evaluate, launch, and transition successful experiments. You might thrive in this role if you: Have a track record of building and shipping end-to-end products in fast-moving, startup, founder-led, growth, or other high-ownership environments. Bring strong frontend engineering skills and enough backend and systems depth to make sound full-stack architectural decisions. Pair product intuition with evidence, using user research and product data to identify opportunities and make pragmatic tradeoffs. Operat

About the Team The Product Explorations team is a small, high-ownership product exploration group focused on turning new model capabilities into impactful user experiences. We operate across Applied and Research, working closely with product, engineering, design, and research partners to identify promising opportunities, prototype quickly, validate ideas, and determine which efforts should scale into larger product investments. Our objective is broad: discover and build experiences that can meaningfully improve user engagement, product impact, and revenue across OpenAI’s products. We move quickly, follow strong product signals, and take ideas from early exploration to meaningful user impact. About the Role We are looking for product-minded full stack engineers to join the Product Explorations team and drive high-impact, net-new product explorations. In this role, you will independently identify promising opportunities, move quickly from idea to prototype, and work across teams to validate and launch new experiences. You will operate with a high degree of autonomy, use strong product intuition, and make thoughtful decisions in ambiguous environments. You may work on a new ChatGPT experience one week, explore a new model capability the next, and partner with Research or Applied teams to turn the strongest ideas into scalable products. This role is a strong fit for someone who enjoys building from scratch, iterating quickly, and working on projects where the path forward is not always obvious. In this role, you will: Identify and pursue high-leverage product and technical opportunities across OpenAI’s products Move quickly from idea to prototype, validation, launch, and iteration Build full stack product experiences that unlock value from new model capabilities Work closely with Research, Applied, product, design, data science, and engineering partners Use user feedback, product signals, and experimentation to determine which ideas should scale Drive alignment across t

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.