AI Operations Enablement Lead

Stark Carpet Corp.; New York (NY); On-site; USD 120,000 - 180,000; Full time

Job summary

STARK Carpet Corp. is building an AI-native operating engine to replace manual processes and link disparate systems. You will map how work actually happens, specify rules, and decide where AI, automation, or software can create value. The goal is to reduce manual effort per order while enabling scalable growth.

Join a high-impact team reporting to the CEO, tackling Finance, Product, Inventory, Warehouse, and HR to turn prototypes into reliable systems and measurements that drive daily decisions.

Responsibilities

• Understand how the work actually happens in Finance, Product, Inventory, Warehouse and HR.
• Map workflows, handoffs, duplications and manual processes.
• Audit existing tools and data flows; identify fragility and owners.
• Translate legacy processes into clear requirements for automation and AI.
• Define where AI, automation, software or process redesign add the most value.
• Own high-value projects (e.g., end-to-end commission reporting).
• Develop dashboards and measurements to monitor ongoing performance.

Job description

Help rebuild how a 90-year-old company operates.

STARK is becoming an AI-native company — not by adding AI on top of broken processes, but by redesigning how work gets done.

We started with one AI Chief of Staff. It worked: internal teams have since eliminated thousands of manual reviews, cut a four-week pricing process to minutes, and returned hours of daily work to Finance.

Now we are building out a three-person AI team reporting to the CEO — one person owning the shared data foundation, one focused on our 200-person sales organization, and this role focused on the operational engine: Finance, Product, Inventory, Warehouse, and HR.

You will walk into $40 million of inventory, millions in freight spend, thousands of manually reviewed orders, legacy systems, inconsistent rules, and leaders asking for these problems to be solved.

Your job is to understand how the work actually happens, simplify it, and build the systems that make the better way stick.

Sometimes that is AI. Sometimes it is software. Sometimes it is automation. Sometimes it is forcing us to finally answer a business question we have avoided answering.

The measure that matters most is effort per order — how much human work it takes to move one order through this company. Lowering it is how we grow without growing headcount.

About STARK

STARK is an 85-year-old luxury carpet and rug company with a strong reputation in the design community: showrooms nationally, a wholesale division, e-commerce, and a warehouse and fabrication hub in Calhoun, Georgia.

We want to feel less like a legacy business and more like a technology company that sells carpet.

Why This Role Exists

Our operating functions are carrying too much manual work, and our systems do not talk to each other well enough for anyone to see the whole picture.

A few real examples:

• One person manually edits every e-commerce order that flows into our ERP
• Our accounts payable team spends three hours a day on manual batch processing
• Commission reporting involves complex rates, account-level overrides, exceptions, and manual judgment — then goes out as roughly 250 individual emails a month
• We carry $40 million in carpet inventory; a 10% improvement in utilization is worth about $4 million — and we do not yet have a metric that would tell us whether we achieved it
• We spent $8.1 million on outbound freight last year with limited analysis by vendor or showroom
• Our order-entry system and ERP can disagree on price: one order read $15,000 in one system and booked at $300,000 in the other

Our people have already proven what is possible: an audit engine that replaced manual review of thousands of orders, a replenishment dashboard, a pricing engine that cut a four-week cycle to minutes, and an AP dashboard that gave a team back three hours a day.

But many of these tools are fragile and maintained by one person.

Part of this job is turning prototypes into systems the company can depend on.

What You’ll Do

Learn how the work actually happens — then make the rules explicit

Before building anything, you will need to understand how the work really happens.

That means:

• Sitting with Finance, Product, and Warehouse teams and watching the work
• Mapping workflows, handoffs, duplicated effort, workarounds, and manual processes
• Auditing the tools already in use — what is live, what is fragile, who maintains it, and what breaks
• Translating messy legacy ways of working into clear requirements
• Identifying where AI, automation, software, or process redesign can create the most value

Some of our hardest problems are not technical.

The rule may live in someone’s head, an old spreadsheet, or a manual override nobody can explain. Automation forces clarity: What is the actual rule? Why does this exception exist? Should it still exist?

We do not want someone who automates the mess.

Help us simplify the mess, then build the system.

Own real, high-value problems

One project you could own is rebuilding commission reporting end to end.

Today that includes rates by product type, customer and salesperson overrides, date-based rules, manual judgment, and exceptions — some documented, some not.

The goal is not to automate the spreadsheet.

The goal is to help Finance and Sales define the rules clearly enough to automate them, build an auditable calculation people can challenge, and reduce Finance’s role to approving genuine exceptions.

Unblock the work we have already committed to

Leadership has active projects across reporting, inventory optimization, warehouse capacity, wholesale planning, partner-location rollout, and financial operations.

Several are waiting on technical work nobody currently owns.

You will help move forward:

• Reporting and dashboards for Product, Finance, and order fulfillment
• Measurement design for efforts that currently have no baseline or metric
• Forecast modeling to shift wholesale reordering from historical to forecast-based
• Integration fixes and cash forecasting on the Finance side
• Exception monitoring and data-integrity checks that surface problems in real time instead of at month-end

Your first job is to make these move.

Help define how an AI-native company actually operates

We are deliberately separating the employee experience layer from our systems of record.

Employees should work through simple, role-specific applications on a shared data foundation, while governed financial transactions remain in core systems.

That creates judgment calls every week:

• AI or