Senior Machine Learning Engineer

kadence; New York (NY); Hybrid; USD 120,000 - 190,000; Full time

Job summary

Kadge Kadence in New York City is seeking a Machine Learning Engineer - MLOps to own production AI systems and keep them reliable as the customer base scales. The role blends infrastructure, evaluation, and monitoring with a strong ownership mindset.

You will collaborate with AI engineers and clinical experts, operate LLM workloads, and build scalable pipelines to track experiments, measure costs, and improve inference performance in a healthcare-focused environment.

Qualifications

• 4 years of experience, incl. 2–3 years in MLOps.
• Strong Python and production-grade LLM workloads.
• Experience owning evaluation for production AI agents.
• Hands-on model development, training and fine-tuning.
• Familiar with MLflow or Weights & Biases for tracking.
• Experience with AWS Bedrock / Azure OpenAI or CI/CD is a plus.

Responsibilities

• Operate production LLM workloads using Kubernetes and orchestrators.
• Build agent evaluation with gates to catch regressions.
• Develop experiment tracking, production monitoring, and versioned pipelines.
• Develop and fine-tune ML models and improve inference performance.

Skills

Python

LLM workloads

Agent evaluation

Experiment tracking

Tools

MLflow

Weights & Biases

CI/CD

AWS Bedrock

Azure OpenAI

Job description

Machine Learning Engineer - MLOps

Hybrid - NYC

We are partnering with a fast-growing healthcare AI startup that recently raised $55M in Series B funding and is scaling AI agents already reviewing real medical claims.

Their mission? Reduce waste in healthcare spending through smarter, more transparent claims operations, and help make care more affordable.

They're hiring an MLOps Engineer to own the infrastructure, evaluation, and monitoring that keep these systems reliable as their customer base grows. You'll report directly to the Head of AI and collaborate with AI engineers, and clinical experts, with meaningful ownership over how the team builds, ships, and measures production AI.

What you'll own:

• Operate production LLM workloads using Temporal or Airflow and Kubernetes, managing retries, timeouts, rate limits, and provider fallbacks.
• Build agent evaluation covering quality, throughput, and cost, with release gates that catch regressions before deployment.
• Develop experiment tracking, production monitoring, and versioned pipelines for models, prompts, and agent configurations.
• Develop and fine-tune ML models and LLMs while improving inference performance and cost.

What they're looking for:

• Around 4 years of experience , including 2-3 years in dedicated MLOps work.
• Strong Python and hands-on experience operating LLM workloads in production.
• Experience owning evaluation for production AI agents.
• Practical model development, training, and fine-tuning experience.
• Familiarity with MLflow, Weights & Biases, or similar tracking and observability tools.

Experience with AWS Bedrock, Azure OpenAI, CI/CD, or healthcare systems would be a plus.

This is a chance to take substantial ownership at a funded startup, work closely with AI leadership, and build systems with a clear real-world purpose.

Applicants must have current U.S. work authorization; visa sponsorship is unavailable.