Senior Manager of Engineering Productivity

Make agentic AI the default way our technology org delivers and operates systems and solutions.

Who We Are:

United States Cold Storage owns and operates one of the most complex temperature-controlled logistics networks in North America. Every day, our systems coordinate the storage and movement of food on a national scale across a network of state-of-the-art distribution centers, including multiple highly automated warehouse facilities.

We continue to advance our core warehouse and logistics platforms. Our current focus is on modular, event-driven, API-first and cloud architecture. We continue to enhance reliability and accelerate engineering productivity by strengthening our SRE and AI practices. This is a large investment in innovation to continue to drive operational excellence at our facilities.

If you want to help us accelerate the building of durable systems that operate in the physical world at scale, this is that opportunity.

The Role:

You will lead the function that makes AI-assisted and agentic tooling (e.g., Cursor, Copilot, Claude Code, MCP-based agents) the default way our technology organization delivers and operates. The bar is not “we turned on Copilot.” The bar is measurable, sustained acceleration across the delivery lifecycle as we modernize mission-critical warehouse, logistics, and corporate systems.

Scope spans requirements through operations: architecture/design, software and data engineering, infrastructure and cloud, customer-facing APIs and EDI integrations, and IT operations. Wherever repeatable toil exists, AI is the lever—and you own the strategy, rollout, and measurable outcomes.

This is not an AI product role. We are not adding AI features to the WMS for customers. We are using AI to build, run, and modernize our technology faster—and at higher quality—than we could without it.

You will partner with functional leaders across Technology, report to the VP of Technology, and have multi-year executive sponsorship. Success is measured in delivery velocity, modernization throughput, defect rate, and reliability—not slideware, pilots, or “AI initiatives launched.”

What You’ll Own

• AI tooling stack — Claude Code, Cursor, Copilot, MCP, agent frameworks. Configuration, fit-for-task playbook, and adoption.
• Productivity baseline — DORA, SPACE, PR cycle time, deployment frequency, MTTR, requirements cycle time, integration time-to-launch. Instrument, measure, move.
• Engineering productivity tooling — internal agents and harnesses for code review, test generation, refactoring, ADR drafting, design critique, and the modernization agent for legacy Java.
• Data, infrastructure, and integration productivity — AI-assisted patterns for pipeline authoring, IaC, observability, and EDI/API integration work (partner spec ingestion, mapping automation, test generation).
• IT operations and agentic ops — agents for change-review, incident triage and remediation, deployment validation, capacity actions, helpdesk knowledge, and runbook automation.
• Modernization playbook — the AI-assisted pattern for safely retiring legacy WMS modules while operations continue.
• Standards, evals, and team — secure usage patterns, prompt/context best practices, model evals, training, and a small senior team to multiply your impact.

What We Are Looking For:

Non-negotiable: you have personally rolled out modern AI coding assistants (e.g., Cursor, Copilot, Claude Code) to a real engineering org (at least dozens of engineers). You can speak concretely about enablement, governance, what you measured, what changed, and what didn’t. You have a defensible view on which tools fit which tasks, and you can show your own setup and workflows.

Beyond that:

• Hands-on builder (platform/automation first). You still get your hands dirty—shipping internal tooling, agents, eval harnesses, and workflow automation (and jumping into code when needed), not just directing others.

• Engineering productivity leader with a measurable operating system. You can instrument the SDLC end-to-end (DORA/SPACE, PR cycle time, requirements lead time, incident/ops metrics), run experiments, and move outcomes quarter over quarter.
• Enablement at scale. You know how to drive adoption across dozens+ engineers with training, champions, playbooks, office hours, and support models—so this sticks beyond early enthusiasts.
• Governance and risk instincts. You build safe-by-default patterns (secure context handling, evaluation and quality gates, auditability) and can partner with Secu