5 months ago
Base Salary
$180k - $440k/yr
Responsibilities
- Design, build, and optimize distributed systems for multimodal pre-training, post-training, inference, data processing, and tokenization at web and petabyte scale.
- Develop high-throughput pipelines for multimodal data acquisition, preprocessing, filtering, generation, decoding, loading, crawling, visualization, and management.
- Advance multimodal understanding and generation, including spatial-temporal compression, cross-modal alignment, world modeling, reasoning, agentic behaviors, and real-time processing.
- Drive human and synthetic data curation, filtering, analysis, and scalable quality pipelines for trillion-parameter models.
- Create evaluation frameworks, internal benchmarks, reward models, and metrics for real-world usage, failure modes, interactive dynamics, and human-AI collaboration.
- Innovate on algorithms, modeling approaches, hardware/software/algorithm co-design, and scaling paradigms.
- Build research tooling, user interfaces, prototypes, demos, full-stack applications, and interactive real-time experiences.
- Work across pre-training, supervised fine-tuning, reinforcement learning, and post-training to enable reasoning, tool calling, agentic behaviors, orchestration, and real-time interaction.
- Collaborate with pre-training, post-training, reasoning, data, applied, and product teams.
Requirements
- Hands-on experience with multimodal pre-training, post-training, or fine-tuning involving vision, audio, video, or cross-modal systems.
- Expert-level Python proficiency and strong experience with at least one of JAX, PyTorch, or XLA.
- A proven track record building or optimizing large-scale distributed machine-learning systems, including training or inference optimization, GPU utilization, multi-GPU or TPU setups, or hardware co-design.
- Deep experience designing and running large-scale data pipelines for curation, filtering, generation, and quality studies involving noisy or real-world multimodal data.
- Strong fundamentals in evaluation design, benchmarks, reward modeling, or reinforcement-learning techniques, particularly for interactive or agentic behaviors.
- Experience leading major improvements in model capabilities through data, modeling, algorithms, or scaling is preferred.
- Familiarity with multimodal LLMs, scaling laws, tokenizers, compression techniques, reasoning, or agentic systems is preferred.
- Rust and/or C++ proficiency for performance-critical components is preferred.
- Experience with large-scale orchestration tools such as Spark, Ray, or Kubernetes is preferred.
- Experience building full-stack tooling, performant interfaces, real-time research demos or applications, or end-to-end products is preferred.
- Strong communication, initiative, prioritization, and ownership are expected.
Benefits
- Equity compensation is included in the total rewards package.
- Comprehensive medical, vision, and dental coverage are provided.
- Access to a 401(k) retirement plan is provided.
- Short- and long-term disability insurance are provided.
- Life insurance and various discounts and perks are available.
Categories
AI ResearchML Engineering
About xAI
Understand the Universe. We are a team of AI technologists and business leaders on a mission to build AI systems that can help humanity understand the world better. https://x.ai/careers