AI Engineer- ML/RL - Weights & Biases
Weights & Biases10 months ago
San Francisco, CA, United States
Hybrid
Full-time
Junior Level (1-3 years)
Job Description
Position Overview
CoreWeave, the AI Hyperscaler™, in partnership with Weights & Biases, is setting a new standard for building, training, and scaling AI. By uniting industry-leading cloud infrastructure with best‐in‐class AI tools, this role empowers you to transform frontier deep learning research into production-grade, reusable solutions for enterprises and the open source community.
Key Responsibilities
- Understand the state-of-the-art in deep learning and translate cutting-edge research into practical, reproducible workflows.
- Build in public by publishing engineering artifacts, reports, and talks that enable others to replicate results.
- Design and ship reference workflows for post-training, agent development, and evaluation using modern frameworks and dashboards.
- Own end-to-end demos covering distributed training, evaluation, and inference—ensuring production-grade quality with clear trade-offs.
- Partner with lighthouse customers, turning recurring patterns into templates and actionable product feedback.
- Track recent advances through papers, releases, and experiments, translating breakthroughs into robust, production-ready solutions.
- Run growth experiments to monitor the usage and impact of the Weights & Biases suite across various use cases.
Required Qualifications
- 5+ years of deep learning experience (training large models in PyTorch or JAX) with strong numerical proficiency.
- Hands-on experience with RL/RLHF frameworks such as SFT, DPO, GRPO, and PPO, including reward modeling and preference-based pipelines.
- Proven production experience in inference and serving models, including familiarity with advanced tools (vLLM, TensorRT-LLM, Triton).
- Expertise in designing evaluation frameworks with statistically sound metrics and robust failure taxonomies.
- Strong proficiency in Python combined with experience in CUDA/Triton kernels, custom C++ operations, or high-performance data ingestion.
- A track record of rigorous experiment tracking and contributions to open source (OSS repos, notebooks, talks) with measurable community adoption.
Preferred Qualifications
- Experience transitioning projects from paper to production in top labs or applied AI teams.
- Familiarity with data engines, feedback loops, and active learning strategies including synthetic data generation.
- Proven success in customer enablement and adoption at scale, demonstrating impact through benchmarked code and results.
Benefits & Perks
- Competitive Salary & Total Rewards: Base pay range from $182,000 to $242,000, with a discretionary bonus, equity awards, and a comprehensive rewards package.
- Medical, dental, and vision insurance – 100% paid for by CoreWeave.
- Company-paid life insurance, plus voluntary supplemental options.
- Short and long-term disability insurance; Flexible Spending and Health Savings Accounts.
- Tuition reimbursement and participation in the Employee Stock Purchase Program (ESPP).
- Mental Wellness Benefits alongside family-forming support, paid parental leave, and flexible childcare options.
- 401(k) with a generous employer match, flexible PTO, and everyday perks like catered lunch and a casual work environment.
Required Skills
Python
Deep Learning
Production Deployment
Benchmarking
Distributed Training
Reinforcement Learning
Experiment Tracking
CUDA
JAX
PyTorch