Deskripsi Pekerjaan
At Nexus Intelligence Labs, we are architecting the next generation of multimodal foundation models that power enterprise automation. We are looking for a Senior Machine Learning Engineer who thrives at the intersection of cutting-edge research and robust, scalable production systems. You will own the full lifecycle of model development—from data curation and distributed training to low-latency inference optimization—directly shaping products used by Fortune 500 companies.
This is a high-autonomy role for a builder who wants to push the boundaries of LLM reasoning, RAG architectures, and agentic workflows. You will work alongside published researchers and world-class engineers in a culture that values intellectual honesty, rapid iteration, and technical excellence.
Why Join Nexus?
- Frontier Tech: Work on 70B+ parameter models, custom CUDA kernels, and massive GPU clusters (H100/A100).
- Ownership: End-to-end ownership of critical model pipelines; your code ships to production weekly.
- Compensation: Top-of-market base, significant early-stage equity (ISO), and 401k matching.
- Environment: Beautiful Financial District office with 3 days/week flexibility, catered lunches, and an annual $3,000 learning stipend.
Tanggung Jawab
- Design, train, and optimize large-scale transformer models (LLMs/VLMs) using PyTorch, FSDP/DeepSpeed, and JAX/Flax on multi-node GPU clusters.
- Architect and maintain high-throughput, low-latency inference infrastructure (TensorRT-LLM, vLLM, Triton) serving millions of daily requests.
- Build robust data flywheels: curate petabyte-scale datasets, implement synthetic data generation pipelines, and design evaluation frameworks for alignment and safety.
- Collaborate with Product and Research to translate novel architectures (Mixture of Experts, State Space Models) from paper to production-grade systems.
- Lead performance profiling and optimization (kernel fusion, quantization INT4/FP8, KV cache management) to reduce cost-per-token.
- Establish MLOps best practices: CI/CD for models, automated drift detection, A/B testing frameworks, and reproducible experiment tracking (Weights & Biases / MLflow).
- Mentor junior engineers and conduct rigorous code/design reviews to maintain a high technical bar.
Kualifikasi
- 5+ years professional experience in Machine Learning Engineering; advanced degree (MS/PhD) in CS, Statistics, or related field preferred.
- Expert proficiency in Python and PyTorch/JAX; deep understanding of GPU architecture, memory hierarchy, and distributed training paradigms (DDP, FSDP, Megatron-LM).
- Proven track record deploying LLMs/VLMs in production: experience with vLLM, TGI, TensorRT-LLM, or custom inference engines.
- Strong systems background: Comfortable with Linux internals, Docker/Kubernetes, CUDA/Triton programming, and profiling tools (Nsight, PySpy).
- Data-centric mindset: Experience with Spark/Ray/Dask for large-scale data processing, data quality auditing, and tokenization strategies.
- Research fluency: Ability to read and implement papers from NeurIPS/ICML/ICLR (e.g., FlashAttention, RoPE, MoE routing, RLHF/DPO).
- Communication & Leadership: Ability to distill complex technical trade-offs for cross-functional stakeholders and drive architectural decisions.