Data Scientist

Vast builds the operating and financial backbone for fast-growing, cash-intensive businesses, combining hands-on execution with purpose-built software, automation, and AI-enabled workflows. We provide technology-enabled shared services, financial infrastructure, and operational support to partners across the U.S. and Puerto Rico, with deep roots in route gaming and other multi-location businesses. We work execution-first, with accurate books, strong controls, and dependable processes, then build the automation and software that raise the standard for how the back office operates.

About the Role

One of the biggest partners we support is a video gaming terminal route across Illinois: machines in bars, restaurants, and truck stops, serviced by field technicians, collected by dedicated crews, supported by a call center, and run on a platform we build and maintain. Every piece of that operation throws off data, and far less of it gets used than should. This role exists to close that gap. Not by producing more charts, but by turning that data into finished work: dashboards that answer a real question, and worklists that tell a specific person what to do Monday morning.

This is not a reporting desk. If the job becomes "run this query for me," we built it wrong. It is not a research role either. Elegance is nice, but a route that runs two hours shorter is better. You will sit close to the operation and to the product and engineering teams building its platform, and your work ships into that live platform, not beside it, and the highest-value work here will be the things nobody thought to request.

What You'll Own

You will work across a deep, multi-year data estate: 250+ Illinois locations, machine and game-level performance, cash and service routing, technician dispatch, the project pipeline, call center volume, and public state reporting. Far more signal than currently gets used.

• Finished analysis, not raw ingredients. A clear answer, the reasoning, and a recommendation someone can act on. Not a table dump.
• Dashboards people open on purpose. Built into the system of record, in our design system, answering questions the regional directors, ops leads, and executives running the route already ask. If nobody opens it twice, it did not work.
• Worklists, the part we care most about. Ranked, assignable lists: the specific machines, locations, or routes that need attention this week, in priority order, with the recommended move and the value of making it. Underperforming machines, wrong collection cadences, equipment to repair or replace, ground lost to nearby competition. A short list, ordered by impact, that an operator can work through.
• Models where they earn their keep. Forecasting, route and schedule optimization, anomaly detection, siting and expected-performance models. Applied, not academic. We care about the decision it changes.

What Success Looks Like

• First 30 days: You know the data model, the metrics, and where the bodies are buried in the data. You have been in the field at least once.
• First 90 days: At least one dashboard and one worklist in real use, with an owner who relies on it.
• First year: Decisions across game mix, routing, staffing, and project prioritization are measurably better because of work you initiated, including work nobody asked for.