Workload Porting And Performance Engineer in San Francisco

About the Team OpenAI’s Infrastructure organization builds and evaluates the systems that power advanced AI workloads. We work closely with hardware, modeling, and architecture teams to ensure that new platforms deliver real-world performance aligned with workload needs. Our team focuses on understanding workload behavior across evolving hardware platforms—bridging the gap between theoretical capability and observed system performance. About the Role We are seeking a Workload Porting & Performance Engineer to evaluate new hardware platforms by porting benchmarks and real-world workloads, analyzing performance, and identifying system bottlenecks. In this role, you will bring up workloads on new systems, characterize performance behavior, and adapt workloads to better utilize hardware capabilities. You will play a critical role in validating new platforms and ensuring that performance aligns with expectations across compute, memory, and networking subsystems. This role requires strong hands-on experience with performance analysis, workload optimization, and system-level debugging across hardware and software boundaries. This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance. Key Responsibilities Port and enable benchmarks and real-world workloads on new hardware platforms. Evaluate system performance across compute, memory, storage, and networking subsystems. Identify and analyze performance bottlenecks and inefficiencies. Adapt and optimize workloads to better utilize hardware capabilities. Develop and run performance experiments and profiling workflows. Compare expected vs. observed performance and provide feedback to: hardware architecture teams performance modeling teams system and software engineers. Debug issues across the stack, including software, runtime, and hardware interactions. Provide actionable insights to guide platform readiness and deployment decisions. Qualifications E

About OpenAI OpenAI is dedicated to ensuring that artificial general intelligence (AGI) benefits all of humanity. Our mission requires building not only world-class AI models, but also the infrastructure that enables those models to be deployed reliably, efficiently, and at global scale. As demand for AI continues to grow, we are expanding the ways OpenAI can bring high-performance inference capacity online across a diverse hardware ecosystem. About the Team The GPT Infrastructure team builds software that turns advanced inference and optimization research into production products. One focus is enabling strategic infrastructure partners and accelerator vendors to qualify and onboard new compute without a bespoke porting and optimization effort for every hardware platform. We build the control planes, APIs, secure partner-side execution environments, evaluation systems, artifact pipelines, and operational tooling that make these workflows repeatable and trustworthy. The work sits at the intersection of distributed systems, AI inference, compilers and runtimes, performance engineering, security, and external partnerships. About the Role We are seeking an experienced systems generalist who can work comfortably across the stack to help build an automated inference optimization platform. Given a workload, target hardware profile, compiler and runtime context, and a trusted verifier, the system runs durable optimization campaigns that generate, compile, execute, grade, and improve candidate kernels, runtime configurations, and serving-stack changes. You will design both the OpenAI-hosted control plane and the partner-side software that evaluates candidates on real accelerator hardware. The product must keep long-running workflows reliable, make performance results reproducible, and maintain clear trust boundaries around sensitive model and hardware information. This is a deeply cross-stack role, combining strong software engineering fundamentals with systems thinking and

About the Team The Scaling team is responsible for the architectural and engineering backbone of OpenAI’s infrastructure. We design and deliver advanced systems that support the deployment and operation of cutting-edge AI models. Our work spans system software, networking, platform architecture, fleet-level monitoring, and performance optimization. About the Role We’re hiring an SW Engineer to enable production workloads and end-to-end testing on new platforms. This role will include creating new test harnesses and platform stress benchmarks, porting existing inference and training workloads to new, sometimes early-access, systems/hardware, analyzing performance and bottlenecks, and characterizing the end-to-end behavior of new systems (compute, comms, storage, control plane, and failure modes). Key Responsibilities Port and validate key inference and training workloads on new platforms/SKUs as they arrive; drive correctness, performance, and stability to an internal readiness bar. Build a suite of benchmarks and stress tests that capture real E2E behavior of our workloads by exercising all aspects of a system, including CPU, GPU, memory subsystem, frontend, scale-up, and scale-out networking (including WAN traffic, NVlink and RDMA collectives), storage, thermals, and any other relevant parts. Deep-dive performance on distributed training/inference: Collective performance and tuning (across NCCL/RCCL and internal libraries) Overlap of compute/communication, kernel-level bottlenecks, memory bandwidth and scheduling effects Create repeatable test harnesses that run in CI / lab environments and produce actionable outputs (pass/fail, performance score, regression detection). Partner with systems + fleet bring-up engineers to ensure the platform is not only stable and performant, but also operationally usable and scalable (containerization, K8s integration, telemetry hooks, failure triage loops). Work cross-functionally with vendors and internal stakeholders by producing

About the Team The Workload Networking team is responsible for the collective communication stack used in our largest training jobs. Using a combination of C++ and CUDA we work on novel collective communication techniques that enable efficient training of our flagship models on our largest custom built supercomputers. The models we train are key ingredients to the AI research progress at OpenAI and the field as a whole, and we continually incorporate learnings from our entire research org into our training platform. About the Role As a Software Engineer, Networking you will design and implement custom networking collectives that are tightly integrated into our training stack. We’re looking for people who have a background in low level performance critical software. Experience with collective communication is a bonus. This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees. In this role, you will: Collaborate closely with ML researchers to design and implement efficient collective operations in C++ and CUDA. Ensure that our largest training jobs take full advantage of the different network transports used in our supercomputers. Work on simulations to inform our future supercomputer network designs. You might thrive in this role if you: Have written distributed algorithms using RDMA in the past. Are comfortable writing low level performance sensitive CPU and/or GPU code. Are familiar with network simulation techniques. About OpenAI OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve our mission, we must encompass and value the many different perspectives, voic

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 We’re hiring software engineers to make the Workload team more productive. The Workload team maintains the core components of OpenAI’s training and inference frameworks and helps execute frontier experiments. About the Role We’re looking for someone who cares about the developer experience of working in and around OpenAI’s core training and inference frameworks. In this role you will: Be responsible for optimizing the development workflows of the engineers around you Work within various Workload teams to address their specific needs, but collaborate with the centralized teams that own various aspects of development experience Optimize iteration speed, both broadly, and in particular by optimizing specific teams’ CI Improve reliability, for instance, by driving testing strategy for particular components Work through the long tail of things that it takes to build libraries and systems that will delight researchers You might thrive in this role if: You are motivated by helping people. You believe a thing that separates great teams from good teams are the players willing to do whatever work it takes, without ego. You believe in the power of developer experience. Something magical happens when people can quickly and confidently iterate on a simple codebase, but this magic is fragile and must be fought for. When you see someone trip over something, no matter how small, your first instinct is asking yourself what it would take for that to not happen again. Your second instinct is clicking merge on the PR you’ve already written to make it so. You are pragmatic. You have the ability to see the world through a perfectionist’s eyes, but are not yourself a perfectionist. You know which problems to pick and when to switch to making progress on a different problem. You like going end-to-end on things. You love co-design — that feeling when you were only able to find the right solution because you both deeply understand the users that interact with a system and the

About the Team The GPT Infrastructure team builds systems that turn advances in model inference and optimization into reliable production capabilities. We enable OpenAI workloads to be qualified and optimized across new accelerator platforms without requiring a one-off port and tuning effort for every hardware target. Our work spans distributed systems, model execution, compilers and runtimes, performance engineering, secure partner integrations, evaluation systems, and developer tooling. We build the infrastructure that makes optimization workflows automated, reproducible, and trustworthy. About the Role We are seeking a software engineer to help build the platform that qualifies and optimizes inference workloads across heterogeneous compute environments. You will develop both OpenAI-hosted services and secure partner-side software for running long-lived optimization workflows. These workflows generate candidate kernels, runtime configurations, and serving-stack changes; compile and execute them on target hardware; verify their correctness; measure their performance; and use the results to guide further optimization. You will work across model architecture, distributed execution, compilers, runtimes, networking, and accelerator systems. A central part of the role is turning research prototypes and one-off hardware bring-up efforts into reliable, reusable infrastructure with clear contracts, reproducible results, strong observability, and well-defined security boundaries. Key Responsibilities Design, build, and operate APIs and control-plane services for long-running workload qualification and optimization campaigns, including scheduling, retries, checkpointing, resource budgets, and observability. Build secure partner-side execution and evaluation software that can compile, run, verify, profile, and benchmark candidate artifacts on accelerator hardware. Integrate model workloads, hardware profiles, compiler toolchains, runtimes, serving engines, and distributed-exe

About the Team OpenAI's Industrial Compute organization builds and operates the infrastructure required to train and serve frontier AI models. The Capacity Planning team connects rapidly changing research and product demand with the compute, networking, storage, power, data center, hardware, and operational resources required to make that demand executable. About the Role We are seeking a Technical Program Manager to build and lead capacity planning across OpenAI's large-scale AI infrastructure. You will translate uncertain workload demand into clear infrastructure requirements, allocation decisions, supply commitments, activation priorities, and long-range capacity strategies. This role sits at the intersection of research, engineering, infrastructure, finance, sourcing, deployment, and operations. You will create the planning models, operating cadences, governance mechanisms, and source-of-truth systems that allow teams to understand what capacity is required, what is available, what is at risk, and what decisions must be made. This is not a finance-only forecasting or reporting role. Success requires technical fluency across the infrastructure stack, strong analytical judgment, and the ability to move consequential decisions forward when requirements, timelines, and supply conditions change quickly. Key Responsibilities Own capacity-planning processes across near-term workload allocation, quarterly execution, and longer-range infrastructure horizons. Translate research, training, inference, and product demand into compute, accelerator, cluster, networking, storage, rack, power, and site requirements. Develop scenarios that make assumptions, confidence levels, constraints, sensitivities, and decision points explicit. Reconcile requested demand against contracted, delivered, installed, activated, and workload-usable capacity. Partner with research and engineering teams to understand workload priorities, technical dependencies, utilization patterns, and changing req

ABOUT BASETEN Baseten powers mission-critical inference for the world's most dynamic AI companies, like Cursor, Notion, OpenEvidence, Abridge, Clay, Gamma and Writer. By uniting applied AI research, flexible infrastructure, and seamless developer tooling, we enable companies operating at the frontier of AI to bring cutting-edge models into production. We're growing quickly and recently raised our $1.5B Series F , led by Altimeter Capital, Conviction Partners, and Spark Capital. Join us and help build the platform engineers turn to to ship AI products. THE ROLE Container runtimes were designed for general-purpose software workloads. AI inference is not a general-purpose workload. Running large models at production scale exposes cracks in every layer of the container stack: runtimes unaware of GPU memory constraints, images that take minutes to pull when a model needs to scale to thousands of replicas, and isolation mechanisms that weren't designed for the multi-tenant serving environments that production AI requires. The tools the industry has relied on for a decade weren't built for this, and patching around those limitations at higher layers only goes so far. Baseten owns the entire pipeline, from the moment a developer pushes a model to the moment a request gets a response. That vertical ownership means we can fix these problems at the root. The Runtime Fabrics team is doing exactly that: purpose-building the container runtime and storage layers for AI inference workloads, led by some of the world's top containerd maintainers. As Engineering Manager of the Runtime Fabrics team, you will lead this work, setting technical direction, growing a world-class team of systems engineers, and ensuring the team's output shapes not just Baseten's infrastructure but the open-source container ecosystem at large. If you've contributed to containerd, runc, or related OCI projects and are ready to lead a team solving some of the hardest problems in infrastructure today, we'd love

About the Team The Storage Infrastructure team builds and operates the storage foundation behind OpenAI’s most demanding workloads. We work directly with research to design storage systems for rapidly evolving experiments, while also powering production at scale. We own the platform end to end: backend systems, user-facing services and APIs, and the control planes that manage how data is placed, moved, and retained over time. Our stack spans cloud and in-house object stores across very different workload profiles, from GPU-attached systems to dedicated storage hardware. We also build the federation layer that unifies these backends behind a simple interface and routes each workload to the right storage solution. About the Role You will help build the storage platform that powers OpenAI’s research and production systems. This is a hands-on infrastructure role for engineers who want to work on deeply technical systems at scale and own them in production. You’ll work across object storage, cross-region data movement, lifecycle management, and the federation layer that provides a unified interface across multiple backends. Much of our stack runs on Kubernetes, and we primarily build services in Rust. In this role, you will: Build and operate storage services that underpin OpenAI’s research infrastructure Develop object storage systems across cloud and in-house environments Build systems for cross-region data movement, replication, and recovery Design lifecycle management capabilities that keep data durable, available, and cost-effective Evolve the federation layer that unifies multiple backend systems behind a simple interface Improve performance, reliability, and operational excellence across the platform Collaborate closely with researchers and infrastructure teams to support rapidly evolving workloads You might thrive in this role if you: Have experience building or operating distributed systems in production Have worked on storage infrastructure, object stores, dist

About the Team At OpenAI, we’re building safe and beneficial artificial general intelligence. We deploy our models through ChatGPT, our APIs, and other cutting-edge products. Behind the scenes, making these systems fast, reliable, and cost-efficient requires world-class infrastructure. The Caching Infrastructure team is responsible for building a caching layer that powers many critical use cases at OpenAI. We aim to provide a high-availability, multi-tenant cache platform that scales automatically with workload, minimizes tail latency, and supports a diverse range of use cases. We’re looking for an experienced engineer to help design and scale this critical infrastructure. The ideal candidate has deep experience in distributed caching systems (e.g., Redis, Memcached), networking fundamentals, and Kubernetes-based service orchestration. In This Role, You Will: Design, build, and operate OpenAI’s multi-tenant caching platform used across inference, identity, quota, and product experiences. Define the long-term vision and roadmap for caching as a core infra capability, balancing performance, durability, and cost. Collaborate with other infra teams (e.g., networking, observability, databases) and product teams to ensure our caching platform meets their needs. You Might Thrive In This Role If You: Have 5+ years of experience building and scaling distributed systems, with a strong focus on caching, load balancing, or storage systems. Have deep expertise with Redis, Memcached, or similar solutions, including clustering, durability configurations, client-side connection patterns, and performance tuning. Have production experience with Kubernetes, service meshes (e.g., Envoy), and autoscaling systems. Think rigorously about latency, reliability, throughput, and cost in designing platform capabilities. Thrive in a fast-paced environment and enjoy balancing pragmatic engineering with long-term technical excellence. About OpenAI OpenAI is an AI research and deployment company d

About the Team OpenAI’s Hardware organization develops system and infrastructure solutions tailored to the demands of advanced AI workloads. We work across the full stack—from silicon to system integration—partnering closely with internal teams and external vendors to define and deliver next-generation AI infrastructure. Our team focuses on defining scalable, high-performance system architectures and reference designs that balance performance, cost, and operational efficiency across rapidly evolving technologies. About the Role We are seeking a 3P Architect to define and drive rack- and cluster-level reference designs in collaboration with external partners. This role is responsible for translating workload requirements and system-level goals into concrete architectures, aligning partners on critical design attributes, and ensuring vendor roadmaps meet our infrastructure needs. You will work closely with performance modeling and internal architecture teams to evaluate tradeoffs, while owning the end-to-end definition and execution of third-party system designs. This includes identifying gaps in current technologies, driving vendor development, and shaping future infrastructure capabilities. This role requires strong system intuition, cross-functional leadership, and the ability to operate effectively across internal teams and external ecosystems. This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance. Key Responsibilities Define rack- and cluster-level reference architectures for AI infrastructure deployments. Translate workload requirements into clear system design specifications and partner deliverables. Collaborate with performance modeling teams to evaluate architectural tradeoffs and system behaviors. Align internal stakeholders and external partners on critical system attributes (performance, cost, power, reliability, scalability). Identify gaps in current technology offerings and dr

About the Team: Compute Infrastructure builds the platform that turns enormous amounts of compute into a reliable engine for frontier AI. We design, provision, schedule, operate, and optimize the systems that connect accelerators, CPUs, networks, storage, data centers, orchestration software, agent infrastructure, developer tools, and observability into one coherent experience for researchers and product teams. Our work spans the entire stack: capacity planning and cluster lifecycle, bare-metal automation, distributed systems, Kubernetes and scheduling, deep system optimization, high-performance networking, storage, fleet health, reliability, workload profiling, benchmarking, and the developer experience that lets teams use enormous compute systems with confidence. At this scale, small improvements to communication, scheduling, hardware efficiency, or debugging workflows can compound into meaningful research velocity. We are hiring across Compute Infrastructure rather than for a single narrow team, and we use this opening to match strong engineers to the problems where they can have the most leverage. About the Role We are looking for engineers who want to build the compute platform behind OpenAI's research and products. You may not be the strongest in low-level systems, high-performance computing, distributed infrastructure, reliability, CaaS, agent infrastructure, developer platforms, tooling, or the user experience around infrastructure. What matters is that you can reason carefully about complex systems, write durable software, and raise the quality and velocity of the people around you. Depending on your background and interests, you might work close to hardware, close to users, on CaaS and agent infrastructure, or on the control planes and data planes in between. You could help bring new supercomputing capacity online, optimize training workloads from profiler traces and benchmarks, improve NCCL and collective communication behavior, reason about GPUs, NICs, t

About the Team OpenAI’s Hardware organization develops system and infrastructure solutions optimized for advanced AI workloads. We collaborate across research, software, and external hardware partners to design and deploy next-generation AI systems at scale. Our team works closely with silicon vendors and system partners to evaluate emerging technologies, validate performance characteristics, and ensure that hardware capabilities translate effectively to real-world AI workloads. About the Role We are seeking a 3P Hardware Architecture Expert with deep expertise in GPU and accelerator architectures to engage directly with silicon vendors and guide hardware decisions for AI infrastructure. In this role, you will evaluate architectural tradeoffs across compute, memory, and interconnect systems, translating vendor specifications into real-world workload impact. You will play a critical role in early silicon evaluation, benchmarking, and performance validation, helping ensure that next-generation hardware meets the needs of our workloads. This role is highly hands-on and requires both deep technical understanding and the ability to engage at a high level with partners such as NVIDIA and AMD on architectural direction and design tradeoffs. This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance. Key Responsibilities Engage deeply with silicon vendors (e.g NVIDIA & AMD) on GPU and accelerator architecture tradeoffs. Analyze and interpret performance, power, and efficiency characteristics of next-generation hardware. Translate vendor specifications into expected real-world performance for AI workloads. Evaluate architectural aspects including: compute throughput and utilization memory systems (HBM, cache hierarchies, bandwidth constraints) data types and precision tradeoffs (FP16, BF16, FP8, etc.) interconnect and scaling behavior. Run benchmarks and profiling to validate hardware performance a

About Pinterest: Millions of people around the world come to our platform to find creative ideas, dream about new possibilities and plan for memories that will last a lifetime. At Pinterest, we’re on a mission to bring everyone the inspiration to create a life they love, and that starts with the people behind the product. Discover a career where you ignite innovation for millions, transform passion into growth opportunities, celebrate each other’s unique experiences and embrace the flexibility to do your best work. Creating a career you love? It’s Possible. At Pinterest, AI isn't just a feature, it's a powerful partner that augments our creativity and amplifies our impact, and we’re looking for candidates who are excited to be a part of that. To get a complete picture of your experience and abilities, we’ll explore your foundational skills and how you collaborate with AI. Through our interview process, what matters most is that you can always explain your approach, showing us not just what you know, but how you think. You can read more about our AI interview philosophy and how we use AI in our recruiting process here . Pinterest is seeking a Sr. Manager to lead our Capacity Engineering team. The team ensures that Pinterest’s cloud infrastructure has the capacity it needs while operating reliably, efficiently and with clear financial accountability. You’ll lead the full portfolio across forecasting and supply, capacity-management systems, compute and GPU efficiency, infrastructure data and governance and capacity operations. What you’ll do: Lead the Capacity Engineering team and establish its 12–18 month functional and technical strategy, roadmap and success measures tied to Infrastructure and company goals. Develop CPU and GPU forecasts and supply plans that account for workload demand, delivery constraints, cost and reliability requirements. Guide the design and delivery of capacity requests, reservations, entitlements, allocation policy and infra