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 to lead the design, training, and deployment of large-scale foundation models that power our enterprise product suite.
You will operate at the intersection of research and production, collaborating with a world-class team of researchers and engineers to push the boundaries of LLM optimization, distributed training, and low-latency inference. This is a high-impact role where your work directly shapes the core intelligence of our platform.
We offer a competitive compensation package including significant equity, comprehensive health benefits, a $3,000 annual learning stipend, and a flexible hybrid work policy anchored in our vibrant Financial District office.
Tanggung Jawab
- Design and implement scalable ML pipelines for pre-training, fine-tuning, and RLHF of Large Language Models (LLMs) on GPU clusters (100+ nodes).
- Optimize model architecture and training objectives for throughput, memory efficiency, and convergence speed using techniques like FlashAttention, FSDP, and Activation Checkpointing.
- Build and maintain high-performance inference serving stacks (TensorRT-LLM, vLLM) targeting sub-100ms latency at scale.
- Collaborate with Research Scientists to reproduce novel architectures from papers (e.g., Mixture of Experts, State Space Models) and transition them into production-grade code.
- Establish rigorous evaluation frameworks (automated benchmarks, human-in-the-loop) to measure model quality, safety, and alignment.
- Mentor junior engineers and define best practices for MLOps, experiment tracking (Weights & Biases/MLflow), and CI/CD for ML artifacts.
- Contribute to open-source tooling and publish findings at top-tier conferences (NeurIPS, ICML, ICLR).
Kualifikasi
- M.S. or Ph.D. in Computer Science, Machine Learning, or a related quantitative field (or equivalent industry experience).
- 5+ years of professional experience building and deploying deep learning models in production environments.
- Expert-level proficiency in PyTorch and the Python data ecosystem (NumPy, Pandas, Hugging Face Transformers/Accelerate).
- Proven track record of distributed training at scale (multi-node, multi-GPU) using NCCL, DeepSpeed, or Megatron-LM.
- Deep understanding of Transformer architecture internals, attention mechanisms, and quantization techniques (GPTQ, AWQ, bitsandbytes).
- Strong software engineering fundamentals: clean code design, Docker/Kubernetes, CI/CD (GitHub Actions/GitLab), and version control.
- Experience with cloud infrastructure (AWS/GCP/Azure), specifically provisioning GPU instances (p4d/p5, A100/H100) and managing storage (S3/GCS) for large datasets.