AI Solutions for Wyoming State Government, Mineral Revenue Forecasting, and F.E. Warren-Adjacent Procurement

ML ensemble models that incorporate commodity price futures, power plant retirement schedules published by the EIA, BLM lease auction results, and Powder River Basin production data from WYOGEO can produce forecasts with materially lower error rates than the regression models Wyoming's Economic Analysis Division has historically used. The 2019–2020 coal revenue shortfall — which created a $400 million budget gap — was partially attributable to forecast models that didn't capture the speed of coal plant retirements. Production-grade ML forecasting for Wyoming mineral revenues typically costs $150,000–$400,000 to build and validate, with annual maintenance running $40,000–$80,000. The ROI threshold is low given the scale of what's being forecasted.

F.E. Warren's contractor community has established FedRAMP and DoD security documentation as informal baseline expectations in Cheyenne's technology procurement culture. State civilian agencies are not legally required to meet military standards, but vendors who arrive without cloud security documentation will face credibility questions from state IT staff accustomed to working alongside defense contractors. For any work that touches F.E. Warren directly — base IT support, contractor supply chain work — formal FedRAMP Moderate or High authorization and DoD Impact Level compliance are required. Wyoming's Office of Enterprise Technology has been moving state systems toward cloud platforms with FedRAMP authorization, which will extend the formal requirement to more state agency applications over time.

Yes — PSC dockets for complex PacifiCorp rate cases can run thousands of pages across hundreds of filings. NLP document classification that automatically tags filings by case, subject matter, and procedural category allows commission staff to locate relevant evidence faster and helps commissioners navigate complex technical proceedings without manually reviewing every document. The PSC's transition planning dockets — covering Rocky Mountain Power's coal retirements and wind integration plans — have been among the most complex in Wyoming regulatory history. AI docket management is a low-risk, high-value application that can be implemented without triggering algorithmic decision-making concerns because it assists human decision-makers rather than replacing them.

Wyoming's Department of Health administers Medicaid on a fee-for-service basis rather than delegating claims adjudication to managed care organizations — one of a small number of states that still do so at this scale. That means the department directly receives and adjudicates individual claims from providers, giving state staff direct access to billing data that MCO states must request from contractors. AI fraud detection models can run directly against the state's claims database in near-real-time. The flip side is that Wyoming doesn't have MCO fraud detection resources to supplement state capacity, so the department's own ML capability matters more. Wyoming's Medicaid population of approximately 80,000 creates a smaller training dataset than most states, which limits model sophistication but also reduces false-positive rates.

UW's Energy Research Center has built ML models for oil and gas production forecasting and wind resource assessment that are directly usable by the Department of Revenue's mineral forecasting function. UW's School of Computing has active ML research that state agencies can access through sponsored research agreements — typically faster and cheaper than commercial vendor engagements for research-stage applications. The Wyoming Technology Business Center in Laramie has incubated AI and analytics firms that have worked on state agency projects. For the Department of Education's school finance modeling and the Department of Environmental Quality's water resource management, UW partnerships provide access to domain expertise and data that commercial vendors would need months to acquire independently.