Senior MLOps Engineer — Real-Time Risk Platform

Mercury; New York (NY); On-site; USD 167,000 - 208,000; Full time

Benefits offered by this job

Equity

Benefits

Salary bands

Job summary

Mercury is expanding its Machine Learning Platform (MLP) to accelerate real-time risk decisions and production observability. The team builds the path from trained models to reliable deployment, delivering low-latency scores to the decision engine and managing end-to-end ML lifecycle.

The ideal candidate will have 5+ years in ML engineering or related fields, strong Python backend skills, and experience deploying models with CI/CD, observability, and drift monitoring.

Qualifications

• 5+ years in machine learning engineering, backend software engineering, MLOps, or related field.
• Production ML service experience: deploying and serving models in low-latency contexts.
• Strong backend fundamentals in Python with API frameworks (FastAPI/Flask).
• Experience with model deployment lifecycles: registries, versioning, CI/CD.
• Observability and alerting for production services; drift is a plus.

Responsibilities

• Build and operate the real-time inference service with low latency and high availability.
• Own model deployment infrastructure: registry, versioning, CI/CD, shadow canary rollouts.
• Build model observability: latency, errors, and drift monitoring.
• Partner with Risk Data Science to move models to production under MLP ownership.
• Implement experimentation like champion/challenger and SHAP explainability outputs.
• Take ownership on small/medium projects and help shape a new platform team.

Skills

5+ years in ML engineering, backend

Production ML service experience

Python backend development

FastAPI/Flask

Model deployment lifecycle tooling

Observability & alerting

SQL & low-latency stores

Streaming pipelines (Kafka/Kinesis)

CI/CD for models

Tools

Python

FastAPI

Flask

CI/CD tooling

Model registries

Shadow/canary rollout

Job description

Mercury is expanding its Machine Learning Platform (MLP) to accelerate real-time risk decisions and production observability. The team builds the path from trained models to reliable deployment, delivering low-latency scores to the decision engine and managing end-to-end ML lifecycle.

The ideal candidate will have 5+ years in ML engineering or related fields, strong Python backend skills, and experience deploying models with CI/CD, observability, and drift monitoring.