Artificial Intelligence Engineer

Stefanini Group • Dearborn, MI

1d • Hybrid

About The Position

Stefanini Group is hiring an Artificial Intelligence Engineer to support AI and ML engineering capability within the TOP platform. This role involves overseeing model fine-tuning, designing agentic orchestration architecture, and evaluating LLMs. The engineer will manage vendor fine-tuning of Google Cloud Vertex AI using proprietary diagnostic data, ensuring compliance with IP protection and model weight storage architecture. The position requires designing and building an Orchestration Layer to integrate external and internal AI engines with TOP platform services. The role also includes evaluating AI engine outputs against accuracy, latency, and first-time fix rate metrics, defining model evaluation frameworks, and building internal tooling for model monitoring and retraining within a GCP environment. Collaboration with data engineering and GCP Cloud Engineers is essential for data preparation, feature engineering, model artifact storage, versioning, and access controls. The AI Engineer will also contribute to the long-term insourcing roadmap by documenting model architectures, training pipelines, and prompt frameworks, and represent AI/ML engineering in architecture reviews and vendor discussions.

Requirements

• 5 or more years of professional experience in machine learning engineering, AI systems development, or applied AI research
• Hands-on experience fine-tuning LLMs in a cloud environment, with specific preference for Google Cloud Vertex AI or equivalent managed ML platforms
• Demonstrated experience building agentic AI systems using frameworks such as LangChain, LangGraph, Google Agent Builder, or equivalent orchestration tooling
• Machine Learning – 3–5 years of applied ML experience including feature engineering, model selection, training, validation, and deployment.
• Candidate should be comfortable working with both structured and unstructured data in the context of real-world engineering or automotive telemetry use cases
• 3–5 years writing production-quality Python for data engineering, ML pipeline development, or platform tooling.
• Proficiency with relevant libraries such as Pandas, NumPy, scikit-learn, and TensorFlow is expected, along with familiarity with code quality practices such as testing and version control.
• 3–5 years of experience designing or working with AI systems, including the application of large language models, expert systems, or intelligent automation within developer or data workflows.
• Candidate should understand model lifecycle management, prompt engineering, and responsible AI practices
• Proficiency in Python and ML development tooling including Hugging Face, PyTorch or TensorFlow, and MLflow or Vertex AI Experiments
• Experience designing and evaluating LLM outputs for production systems, including prompt engineering, retrieval-augmented generation (RAG) architectures, and model evaluation metrics
• Strong understanding of MLOps practices including model versioning, deployment pipelines, monitoring, and retraining workflows on GCP
• Experience working in regulated or IP-sensitive environments where model artifact ownership and data governance are active concerns
• Google Cloud Platform – 2–5 years of hands-on experience with GCP services relevant to AI/ML and data workloads, including Vertex AI, BigQuery, GCS, Dataflow, or Cloud Composer, with the ability to deploy and manage workloads in a production cloud environment

Nice To Haves

• Experience in automotive diagnostics, vehicle telematics, or connected vehicle platforms.
• Familiarity with Diagnostic Trouble Code (DTC) data, Over-the-Air (OTA) update systems, or repair order (RO) data structures
• Experience with multi-agent AI systems and tool-use patterns in production
• Google Cloud Professional Machine Learning Engineer certification

Responsibilities

• Design and build Orchestration Layer.
• The integration framework that connects external AI engine with other internal AI engines and TOP platform services
• Evaluate AI engine outputs against defined accuracy, latency, and first-time fix rate metrics; drive iterative improvement through structured feedback loops
• Define model evaluation frameworks and acceptance criteria for AI-generated triage recommendations, ensuring clinical accuracy before dealer-facing deployment
• Build internal tooling for model monitoring, drift detection, and retraining triggers within GCP environment
• Collaborate with data engineering team to define data preparation and feature engineering requirements that support model fine-tuning and inference quality
• Partner with the GCP Cloud Engineers to ensure model artifact storage, versioning, and access controls comply with IP and security policies
• Contribute to the long-term insourcing roadmap by documenting model architectures, training pipelines, and prompt frameworks in sufficient detail to enable internal replication
• Represent AI and ML engineering in architecture reviews and vendor technical discussions.

Benefits

• Listed salary ranges may vary based on experience, qualifications, and local market.
• some positions may include bonuses or other incentives

What This Job Offers

• Job Type: Full-time
• Career Level: Senior