AI/ML Engineer 2

Role overview

AI/ML Engineer 2 at Day & Zimmermann. You will own the end-to-end design, development, and deployment of advanced AI/ML solutions supporting D&Z AI initiatives, including generative AI (LLM agents, RAG systems, MCP servers) and predictive ML models.

Responsibilities

• Design, build, and deploy production-grade generative AI applications, including:
  • AI agents
  • RAG chatbots
  • workflow automations
  • MCP servers
  • using modern frameworks and vector databases
• Design, develop, and deploy predictive models using ML techniques across:
  • regression, classification, clustering, and time series
• Author custom skills, prompts, slash commands, and evaluation harnesses; use AI coding tools such as Claude Code, Codex, and Cursor
• Own production monitoring, including drift detection, online/offline metrics, and retraining strategies
• Own data engineering and pipeline infrastructure for AI/ML systems, including:
  • retrieval architectures
  • embeddings
  • context engineering
  • training/fine-tuning workflows
• Review code, mentor junior engineers, and contribute to engineering standards and documentation
• Present results and deliver live demos to technical and executive audiences

Requirements (KSAs)

• Deep hands-on experience building production LLM applications with providers such as:
  • Anthropic, OpenAI, Google, and open-source models
• Strong prompt engineering and context engineering; expertise in:
  • RAG with vector databases
  • AI agent design and tool use
  • Model Context Protocol (MCP) server development
• Experience authoring custom skills and building evaluation harnesses
• Proficiency with ML frameworks including scikit-learn, XGBoost, PyTorch, TensorFlow
• Strong foundation in statistical analysis, hypothesis testing, feature engineering, model optimization, and monitoring/deployment
• Ability to design/validate experiments and define online/offline metrics using quantitative methods
• Strong proficiency in Python for AI/ML application development

Nice-to-haves

• Comfort communicating complex technical concepts to both technical and non-technical audiences
• Experience with online/offline evaluation and iterative improvement loops in production