Deskripsi Pekerjaan
Shape the Future of Generative AI
Nexus Intelligence Labs is seeking a Senior Machine Learning Engineer to architect and deploy the next generation of large-scale foundation models. We are a well-funded, stealth-mode startup backed by top-tier VCs (Sequoia, a16z), pushing the boundaries of multimodal reasoning and agentic workflows.
You will own the full lifecycle of model development—from data curation and distributed training on massive GPU clusters to low-latency inference optimization and production monitoring. This is a high-autonomy role for a builder who thrives in ambiguity and wants to define how AI interacts with the world.
Why Join Us?
- Technical Ownership: End-to-end ownership of critical model pipelines.
- Compute Budget: Unrestricted access to 10,000+ H100 GPUs.
- Open Culture: Publish at NeurIPS/ICML/ICLR; contribute to open source.
- Compensation: Top-of-market salary, significant equity (ISO), 401k matching, and comprehensive healthcare.
If you are ready to solve the hardest problems in AI with a world-class team, apply today.
Tanggung Jawab
- Architect & Train: Design, implement, and scale novel model architectures (Transformers, MoE, SSMs) for multimodal understanding and generation using PyTorch/JAX.
- Distributed Systems: Optimize training throughput on large-scale clusters (1000+ GPUs) using FSDP, DeepSpeed, Megatron-LM, and custom kernel fusion (Triton/CUDA).
- Data Centric AI: Build automated data curation, synthesis, and evaluation pipelines to continuously improve dataset quality at petabyte scale.
- Inference Engineering: Deploy models with sub-100ms latency using TensorRT-LLM, vLLM, or custom inference stacks; implement quantization (GPTQ, AWQ), speculation, and KV-cache optimization.
- R&D Leadership: Stay current with literature; reproduce and extend SOTA papers; drive research direction for pre-training, alignment (RLHF/DPO/RLAIF), and agentic tool use.
- Mentorship: Mentor junior engineers; establish coding standards, design review processes, and testing rigor for ML systems.
- Cross-functional Collab: Partner with Product and Research to translate business requirements into technical specifications and measurable model metrics.
Kualifikasi
- Education: MS/PhD in Computer Science, Machine Learning, Mathematics, or equivalent practical experience (5+ years industry ML experience).
- Framework Mastery: Expert-level proficiency in PyTorch (required) or JAX; deep understanding of autograd, compilation (torch.compile/XLA), and memory management.
- Large Scale Training: Proven track record of training models >10B parameters on multi-node clusters; hands-on experience with NCCL, RCCL, and cluster orchestration (Slurm/Kubernetes).
- Systems Programming: Strong software engineering fundamentals in Python and C++/Rust; experience writing high-performance kernels (CUDA/Triton) is a major plus.
- LLM Expertise: Deep practical knowledge of LLM pre-training, instruction tuning, alignment techniques (RLHF, DPO, ORPO), and evaluation methodologies (MMLU, MT-Bench, custom evals).
- MLOps & Tooling: Experience with MLflow, Weights & Biases, Ray, Kubeflow, or internal platform tooling for experiment tracking and reproducibility.
- Communication: Ability to articulate complex technical concepts to both technical and non-technical stakeholders; strong technical writing skills.