Machine Learning Engineer

Details:

Stefanini Group is hiring!

Stefanini is looking for Machine Learning Engineer, Dearborn, MI

For quick apply, please reach out to Lovkesh Bharti at 248-727-2506/lovkesh.bharti@stefanini.com

We are seeking a high-impact AI/ML Engineer to build intelligent data products that turn complex, high-volume engineering information into trusted, actionable insight. You will work across applied machine learning, generative AI, data platforms, and cloud engineering to deliver production systems used for search, traceability, analytics, and decision support. This role is ideal for an engineer who can move from architecture to implementation to operational ownership, and who enjoys solving ambiguous problems where data quality, scale, and reliability matter.

Responsibilities

• Architect, build, and operate reliable data products that ingest and transform diverse structured and unstructured information at enterprise scale.
• Create resilient orchestration and delivery patterns for batch and near-real-time workloads, with clear observability, alerting, and operational runbooks.
• Develop production Retrieval-Augmented Generation (RAG) systems that combine semantic retrieval, structured data, and grounded responses for high-value engineering use cases.
• Design agentic AI workflows that decompose complex questions, select the right data sources and tools, validate results, and return explainable answers with citations.
• Develop and evaluate embedding, document-understanding, and multimodal inference workflows, balancing quality, latency, scalability, and cost.
• Lead cloud architecture, containerization, infrastructure-as-code, and CI/CD practices for secure, repeatable deployment across environments.
• Own system reliability from design through production: investigate incidents, profile performance, eliminate failure modes, and improve capacity planning.
• Deliver intuitive analytics experiences and decision-support tools that make complex technical data useful to engineers, program teams, and leadership.
• Establish data quality, lineage, validation, and governance practices so users can understand where information came from and how much to trust it.
• Build incremental, restartable processing with checkpointing and recovery strategies that protect data integrity during long-running or partially failed workloads.