Deskripsi Pekerjaan
Nexus Intelligence is building the decision layer for the modern enterprise. We sit at the intersection of Large Language Models, causal inference, and high-frequency data streaming. We are looking for a Senior Machine Learning Engineer who thrives in ambiguity, treats ML as an engineering discipline, and wants to ship production-grade AI that impacts Fortune 500 bottom lines.
This is not a research role—it is a build and scale role. You will own the full lifecycle: from data architecture and feature stores to model serving infrastructure and rigorous A/B testing frameworks. You will work alongside ex-FAANG engineers and published researchers in a high-trust, low-bureaucracy environment.
Why This Role Is Different
- Real Production Scale: Models serving millions of predictions/day with sub-50ms latency SLAs.
- Modern Stack: Python, PyTorch, JAX, Ray, Kubernetes, Dagster, Weights & Biases, FeatureForm.
- Ownership: End-to-end responsibility—no "throw it over the wall" to ML Ops.
- Compensation: Top-of-market base, significant early-stage equity, and a 401k match.
Tanggung Jawab
- Design, build, and maintain scalable ML pipelines for training, evaluation, and deployment of LLMs and classical ML models.
- Architect robust feature engineering systems and feature stores supporting both batch and real-time inference.
- Optimize model serving infrastructure for latency, throughput, and cost-efficiency using TorchServe, Triton, or custom containers on K8s.
- Establish rigorous offline/online evaluation frameworks, including shadow testing, A/B testing, and drift detection.
- Collaborate with Product and Data Engineering to define data contracts, schema evolution strategies, and labeling workflows.
- Implement MLOps best practices: CI/CD for ML (CT), automated retraining triggers, and model registry governance.
- Mentor junior engineers and conduct code/design reviews upholding high engineering standards.
- Stay current with SOTA architectures (RAG, Agents, Mixture of Experts) and assess applicability to our roadmap.
Kualifikasi
- 5+ years of professional experience building and deploying ML systems in production environments (not just notebooks).
- Expert-level Python; strong proficiency in at least one systems language (Go, Rust, C++).
- Deep hands-on experience with PyTorch or JAX; TensorFlow experience a plus.
- Proven track record designing ML infrastructure: Feature Stores, Model Registries, Orchestration (Airflow/Dagster/Prefect).
- Strong software engineering fundamentals: Clean architecture, testing (unit/integration/contract), CI/CD, GitOps.
- Experience with cloud-native stacks (AWS/GCP/Azure), Kubernetes, and Infrastructure-as-Code (Terraform/Pulumi).
- Demonstrated ability to debug complex distributed systems and optimize GPU utilization.
- BS/MS/PhD in Computer Science, Statistics, or equivalent practical experience.