AI Engineer

Sunset•Livonia, MI

2d•Onsite

About The Position

Mastronardi Produce is seeking an AI Engineer for its corporate office in Livonia, Michigan. This role is responsible for building, deploying, and maintaining AI-powered applications and integrations that extend the company's AI Enablement platform. The focus is on transforming AI capabilities, including large language models, retrieval, and agentic workflows, into production-ready tools that are reliable, governed, and integrated into business systems and processes. This is a hands-on individual contributor role where the engineer will write code, build integrations, and deliver working solutions, taking accountability for the technical architecture, performance, and security of deployed solutions. The role requires adherence to Mastronardi's PRIDE values: Passion, Respect, Innovation, Drive, and Excellence.

Requirements

At least 5 years experience in AI Engineering
Diploma/Degree in related discipline
Software development experience in Python (required) and at least one other language (e.g., JavaScript/TypeScript, C#, or similar).
Hands-on experience building applications with large language models — prompt engineering, RAG, tool/function calling, and agentic workflows.
Experience with LLM platforms and APIs such as Claude (Anthropic), Azure OpenAI, or equivalent.
Experience building and consuming REST APIs and integrating disparate systems.
Working knowledge of vector databases, embeddings, and retrieval techniques.
SQL: ability to query, join, and filter data across relational databases.
Experience with cloud platforms (Microsoft Azure preferred) including AI/ML services such as Azure AI Foundry.
Version control and CI/CD practices (Git, automated testing, deployment pipelines).
Understanding of AI safety, governance, and evaluation practices — ensuring outputs are auditable and appropriate for business use.

Nice To Haves

Familiarity with ERP or enterprise data structures (Microsoft Dynamics NAV/365, SAP, or equivalent) is an asset.
Understanding of responsible AI practices, including bias evaluation and hallucination mitigation, to keep model outputs safe and appropriate for business use.
Experience with containerization and deployment tooling (e.g., Docker, Azure App Service or Functions) is an asset.

Responsibilities

Design, build, and deploy AI-powered applications and integrations that connect large language models (e.g., Claude, Azure OpenAI) to business systems and data sources.
Build retrieval-augmented generation (RAG) pipelines and knowledge bases that ground AI outputs in accurate, current company data.
Develop and refine prompts, system instructions, and evaluation frameworks to ensure AI outputs are accurate, consistent, and auditable.
Build the initial APIs, connectors, and integration logic between AI platforms and enterprise systems (ERP, WMS, data platforms, Microsoft 365) as part of solution design, handing off production deployment and ongoing operation to Cloud Operations & Infrastructure.
Define the guardrails, audit points, and monitoring requirements AI solutions need so outputs can be reviewed, debugged, and improved after deployment, partnering with Cloud Operations & Infrastructure on the monitoring infrastructure that implements them.
Test AI solutions rigorously before release, including functional testing, edge-case validation, and structured evaluation against real business scenarios.
Partner with the Data Engineering and Data Governance teams to ensure AI solutions use well-structured, quality data and comply with security and privacy requirements.
Stay current on the AI tooling landscape (models, frameworks, agent platforms) and recommend where new capabilities apply to real business problems.
Build reusable components, templates, and internal libraries that speed up delivery of future AI solutions.
Create and maintain clear technical documentation for all solutions, architectures, and operating procedures.
Ensure all solutions align with IT governance, security, and compliance standards.
Design and maintain evaluation suites and benchmarks to track model and prompt performance over time, catching regressions before they reach production.
Contribute to architecture decisions on model selection and hosting (e.g., Azure AI Foundry vs. direct API integration), weighing cost, performance, and reliability tradeoffs.
Support incident response for AI-related production issues and participate in post-incident reviews to identify root cause and prevent recurrence.