Staff Machine Learning Engineer, Platform (MLOPS)
Credit Acceptance1 day ago
Southfield, MI, USA +2 moreStaff+
Base Salary
$185k - $259k/yr
Responsibilities
- Own training and inference pipelines, model registry and versioning, model serving endpoints, and controlled promotion across development, QA, and production.
- Operate production model runtimes through monitoring, alerting, drift and quality-regression detection, latency and throughput objectives, autoscaling, and incident response.
- Operate agent runtimes through the enterprise AI Gateway and MCP Gateway, including governed routing, tool scoping, versioning, authorization, and migration of existing agents.
- Build production agent evaluation, online scoring, behavioral monitoring, trace and telemetry capture, sampling and judge pipelines, and regression-blocking release gates.
- Deliver reusable pipeline templates, deployment patterns, reference implementations, internal tooling, observability infrastructure, and evaluation data contracts.
- Measure inference, document, and interaction costs and provide platform and infrastructure analysis for build-versus-buy and hosting decisions.
- Partner with Cloud Engineering, Data Engineering, Security, and SRE on enterprise governance, identity, observability, and AI-specific incident response.
- Maintain ML platform architecture documentation and mentor engineers and interns through design and code reviews.
Requirements
- Bachelor's degree in Computer Science, Engineering, Statistics, or a relevant technical field and at least 7 years of relevant experience, or a master's degree in one of those fields and at least 5 years of relevant experience.
- At least 5 years building and operating production ML or AI systems, with direct ownership of at least two major platform areas such as pipelines, model serving, registries, or production monitoring.
- Experience operating a production ML or AI service through deployment, paging, failure diagnosis, and corrective fixes.
- Strong Python and SQL skills with production engineering practices applied to ML workloads.
- Production experience with a cloud ML platform; AWS and Databricks are strongly preferred, including model serving, job orchestration, and MLflow or a similar registry or experiment-tracking system.
- Working knowledge of LLM and GenAI operational concerns, including token cost, latency, caching, batching, and nondeterministic output.
- Experience operating LLM or agent applications behind gateway or proxy layers, including model routing, fallback, credential management, rate limiting, and budget enforcement.
- Knowledge of tool-calling architecture and the Model Context Protocol, including MCP server operation, tool authorization, and gateway-based access.
- Experience with containerization and infrastructure as code.
- Preferred experience includes GPU-backed model serving, OpenTelemetry, Dynatrace, model-serving optimization, Databricks Unity Catalog, regulated-industry AI operations, agentic systems, managed agent or tool gateways, MCP servers, and Agent2Agent interoperability standards.
- Ability to communicate technical trade-offs clearly, resolve incidents, prioritize under pressure, teach others, and build cross-functional relationships.
- Must reside in or be willing to relocate to the greater Southfield, Michigan area; greater Bellevue, Washington area; or San Francisco Bay Area, California.
Benefits
- Competitive base salary of $184,943–$258,920 plus eligibility for a 10–20% annual variable incentive of cash and equity.
- 401(k) match, adoption assistance, parental leave, tuition reimbursement, and comprehensive medical, dental, and vision coverage.
- Hybrid work model with two work-from-home days and three standard in-office days each week.
- Additional nonstandard benefits and an inclusive, equal-opportunity workplace.
Categories
About Credit Acceptance
Credit Acceptance is a public auto finance company that partners with a nationwide network of car dealers to provide indirect vehicle loans to consumers, including those with limited or challenged credit. The company purchases and services retail installment contracts and shares collections with dealers through structured programs. Founded in 1972 and headquartered in Southfield, Michigan, it trades on NASDAQ as CACC and operates across the U.S.