Deskripsi Pekerjaan
Shape the Future of Generative AI
Nexus Intelligence Labs is seeking a Senior Machine Learning Engineer to architect and deploy state-of-the-art foundation models that power our next-generation enterprise platform. You will operate at the intersection of research and production, turning novel architectures into scalable, reliable systems used by Fortune 500 companies.
This is a high-autonomy role for a builder who thrives in ambiguity and possesses deep expertise in distributed training, LLM optimization (quantization, distillation, speculative decoding), and MLOps best practices. You will own the full lifecycle: from data curation and model architecture design to serving infrastructure and rigorous evaluation.
Why Join Nexus?
- Compute Access: Dedicated access to large-scale H100/A100 clusters (no queueing).
- Research Freedom: 20% time allocated for exploratory research & publications.
- Competitive Equity: Early-stage equity package with quarterly liquidity events.
- Modern Stack: PyTorch, JAX, Triton, Kubernetes, Ray, Dagster, Weights & Biases.
Tanggung Jawab
- Design, implement, and optimize large-scale distributed training pipelines for LLMs and multimodal models (10B+ parameters) across thousands of GPUs.
- Lead model compression and inference optimization initiatives: quantization (GPTQ, AWQ), knowledge distillation, speculative decoding, and custom kernel development (Triton/CUDA).
- Architect robust evaluation frameworks (automated + human-in-the-loop) for alignment, reasoning, and safety benchmarks.
- Build and maintain the core MLOps platform: feature stores, experiment tracking, CI/CD for models, and high-throughput/low-latency serving infrastructure (vLLM/TGI).
- Collaborate with Product and Research teams to define data strategies: synthetic data generation, curriculum learning, and RLHF/RLAIF pipelines.
- Mentor junior engineers, conduct rigorous code reviews, and establish technical standards for reproducibility and testing.
- Publish findings at top-tier venues (NeurIPS, ICML, ICLR) and represent Nexus at industry conferences.
Kualifikasi
- Education: MS/PhD in Computer Science, Machine Learning, or equivalent practical experience (5+ years industry ML).
- Framework Mastery: Expert-level PyTorch/JAX; proven experience writing custom training loops, distributed strategies (FSDP, DeepSpeed, Megatron-LM), and custom kernels.
- LLM Expertise: Hands-on experience pre-training or fine-tuning LLMs > 7B params; deep understanding of Transformer architecture, attention mechanisms, and scaling laws.
- Systems & Infra: Strong software engineering fundamentals (Python, Rust/Go/C++); proficiency with Kubernetes, Docker, and cloud providers (AWS/GCP/Azure).
- Inference Optimization: Track record of reducing latency/cost for production LLM serving (experience with vLLM, TensorRT-LLM, Tensor Parallelism).
- Data-Centric AI: Experience building large-scale data processing pipelines (Spark/Ray/DataBricks) for curation, deduplication, and quality filtering.
- Communication: Ability to distill complex technical concepts for cross-functional stakeholders and write clear design docs.