Deskripsi Pekerjaan
At Nexus Intelligence Labs, we are architecting the next generation of generative AI infrastructure. We are looking for a Senior Machine Learning Engineer who thrives at the intersection of research and production engineering. You will own the lifecycle of large-scale models—from data curation and distributed training to low-latency serving and rigorous evaluation.
This is a high-impact role where your work directly shapes core product capabilities used by millions. You will collaborate with a world-class team of researchers and engineers in a culture that values intellectual honesty, rapid iteration, and technical excellence.
Why Join Us?
- Frontier Tech: Work with H100 clusters, custom CUDA kernels, and the latest attention architectures.
- Ownership: End-to-end responsibility for model performance, not just ticket-taking.
- Compensation: Top-of-market base, significant equity grant, and comprehensive healthcare/wellness packages.
- Environment: Beautiful SoMa office with 3 days/week flexibility, catered lunches, and an annual L&D stipend.
Tanggung Jawab
- Design, implement, and optimize distributed training pipelines for LLMs and multimodal models on GPU clusters (PyTorch, FSDP, DeepSpeed, Megatron-LM).
- Build and maintain robust MLOps infrastructure: feature stores, experiment tracking (MLflow/W&B), CI/CD for model deployment, and automated evaluation harnesses.
- Collaborate with Research Scientists to translate novel architectures (MoE, RWKV, RetNet) into production-ready code.
- Optimize inference latency and throughput using TensorRT-LLM, vLLM, or custom kernel fusion techniques.
- Curate and process massive datasets (petabyte-scale) using Spark/Ray/DataBricks for pre-training and instruction tuning.
- Establish rigorous offline/online evaluation metrics (human preference alignment, hallucination rates, latency percentiles).
- Mentor junior engineers and define technical standards for code quality, reproducibility, and system observability.
- Participate in on-call rotation for critical model serving infrastructure (Tier 1 latency SLAs).
Kualifikasi
- Education: MS/PhD in Computer Science, Machine Learning, or equivalent practical experience (5+ years industry ML engineering).
- Deep Learning Expertise: Proven track record training and deploying Transformer-based architectures (BERT, GPT, LLaMA, Diffusion) at scale (>10B params).
- Systems Fluency: Expert-level Python/C++; deep understanding of GPU architecture, memory management, CUDA, NCCL, and distributed systems primitives.
- Framework Mastery: Extensive production experience with PyTorch (torch.compile, DTensor), JAX/Flax, or TensorFlow.
- MLOps & Data: Hands-on experience with Kubernetes (Kubeflow/KServe), Airflow/Dagster, and large-scale data processing (Spark, Polars, Ray Data).
- Optimization: Demonstrable experience with quantization (GPTQ, AWQ, FP8), distillation, speculative decoding, or custom Triton kernels.
- Communication: Ability to articulate complex technical trade-offs to both technical and non-technical stakeholders.
- Preferred: Publications at top-tier venues (NeurIPS, ICML, ICLR, MLSys) or significant open-source contributions to ML infrastructure projects.