兼职 MLOps Engineer - LLMOps Engineer - Model Deployment Engineer 工作, 薪水, Psychology Internships 公司招聘中 - Ricebowl

Role Description

Build and operate the systems that deploy, monitor, and maintain machine learning and large language models in production. Partner with data scientists, researchers, software engineers, and security teams to deliver reliable, scalable, and cost-effective AI services.

Key Responsibilities

• Develop automated pipelines for model validation, deployment, versioning, and rollback.
• Build and maintain model serving infrastructure for batch, real-time, or streaming workloads.
• Establish CI/CD practices for model code, configurations, dependencies, and supporting infrastructure.
• Maintain traceability across datasets, experiments, model artifacts, and production releases.
• Monitor service availability, latency, throughput, resource utilization, and model performance.
• Detect data quality issues, drift, and performance degradation, coordinating investigation and remediation.
• Optimize inference performance, compute capacity, and operating costs.
• Implement deployment safeguards, including staged releases, access controls, secrets management, and audit logging.
• Coordinate retraining, model updates, and retirement according to defined validation and approval processes.
• For LLM systems, manage prompt versions, model configurations, evaluation datasets, and application dependencies.
• Support retrieval pipelines, embedding updates, and vector indexes where applicable.
• Evaluate LLM application quality, safety, and reliability, including failures involving retrieved content or tool use.
• Maintain operational documentation, incident response procedures, and recovery plans.

Qualifications

• Relevant experience in MLOps, platform engineering, DevOps, machine learning engineering, or model deployment.
• Proficiency in Python and familiarity with software engineering practices, APIs, and automated testing.
• Experience with containers, orchestration, CI/CD, and infrastructure as code.
• Working knowledge of cloud platforms and production monitoring tools.
• Understanding of the machine learning lifecycle, including training, evaluation, deployment, and monitoring.
• Experience deploying models through inference services or managed model platforms.
• Familiarity with model registries, experiment tracking, and artifact management.
• Strong troubleshooting skills across application, infrastructure, and data workflows.
• Ability to collaborate across technical teams and communicate operational risks clearly.
• A degree in computer science, engineering, data science, or a related field, or equivalent practical experience.

Preferred Qualifications

• Experience with distributed inference, accelerator hardware, model quantization, or inference batching.
• Familiarity with LLM APIs, self-hosted language models, retrieval-augmented generation, and evaluation pipelines.
• Experience managing privacy, security, and governance requirements for production AI systems.
• Knowledge of workflow orchestration, feature pipelines, or feature stores.
• Experience supporting incident response and service reliability targets.

Scope by Role

• MLOps Engineer: Owns infrastructure and automation across the machine learning lifecycle, including training pipelines, model registries, deployment, and monitoring.
• LLMOps Engineer: Focuses on operating LLM applications, including prompt and configuration management, retrieval dependencies, evaluations, observability, and inference costs.
• Model Deployment Engineer: Specializes in packaging, serving, integrating, and optimizing models for target production environments.