Loading...
Loading...
Cheyenne has become one of the most consequential data-center build-out cities in the Mountain West over the last fifteen years, and that single fact reshapes the metro's ML demand profile beyond what the population suggests. Microsoft's North Bay campus on Bishop Boulevard, Meta's data center on Christensen Road south of the city, NCAR-Wyoming Supercomputing Center north of Cheyenne in the Cheyenne Business Parkway area (and its Frontera, Cheyenne, and Derecho HPC heritage), and the smaller hyperscale and colocation footprints across the Cheyenne Logistics Park and Bison Business Park drive a kind of ML demand that no other Wyoming city sees. The work clusters around hyperscale capacity planning, power and cooling forecasting against the unusually cold and dry High Plains climate (which makes Cheyenne attractive for data centers in the first place), grid-interconnection modeling with Black Hills Energy and the WAPA western transmission system, and adjacent ML demand from the F.E. Warren Air Force Base ICBM mission and the state-government data infrastructure operated out of the Capitol complex on Capitol Avenue. The University of Wyoming's School of Computing in Laramie, plus the NCAR-Wyoming partnership for atmospheric and earth-science HPC, supply most of the senior ML talent that lands in Cheyenne. LCCC (Laramie County Community College) feeds the data center technician and operations-analyst tier. LocalAISource matches Cheyenne operators with ML practitioners who have shipped hyperscale capacity, energy-ML, or HPC-adjacent work — three patterns that intersect almost uniquely here.
Updated May 2026
The Microsoft North Bay campus and Meta's Cheyenne facility together represent more than a billion dollars of cumulative investment, and the ML demand they drive is mostly internal to the parent companies. Microsoft's Azure capacity-planning organization, Meta's data-center-services group, and the shared infrastructure-engineering teams that span both companies run sophisticated ML for thermal modeling, power-usage-effectiveness optimization, server-fleet failure prediction, and capacity-versus-utilization forecasting. Most of that work runs centrally, not through Cheyenne procurement. Where outside ML partners land work is in the supporting infrastructure — Black Hills Energy's grid-side modeling for the data center load, WAPA-related transmission capacity work, and the colocation and Tier 2 cloud operators in the Cheyenne Logistics Park whose ML programs are smaller and more accessible. Capacity-planning ML for these mid-tier operators covers thermal-load forecasting against High Plains weather (which has unusually predictable cold-and-dry patterns useful for free-cooling), generator and UPS predictive maintenance, network-traffic forecasting, and SLA-margin modeling. Engagement budgets in the colocation tier run sixty to two-fifty thousand. The hyperscale-adjacent work — particularly with Black Hills Energy on the grid side — runs larger, into the four-hundred-thousand range when serious load-forecasting and demand-response modeling is in scope. ML partners with prior data-center-engineering or grid-side experience are the only realistic candidates.
The data-center concentration in Cheyenne has reshaped Black Hills Energy's local distribution and transmission planning, and the utility runs ML demand that did not exist at the same scale a decade ago. Load forecasting against hyperscale customers requires fundamentally different modeling than residential or commercial load — hyperscale loads are large, sudden, and tied to compute demand patterns that include weekday-versus-weekend variation, AI-training-job batching, and step-function changes when a new building energizes. Demand-response ML for the colocation operators that run on backup generators or behind-the-meter assets, integration ML for the WAPA western transmission system that ties Cheyenne loads into the broader western grid, and rate-design and tariff modeling all create engagement opportunities. Beyond Black Hills Energy directly, Wyoming Public Service Commission filings often involve modeling work that supports rate cases and grid-interconnection decisions. The University of Wyoming's School of Energy Resources and the Wyoming Energy Authority occasionally fund collaborative ML work that touches grid-side problems. Engagement budgets here run one-fifty to four-hundred-thousand for serious utility ML, and the work pulls heavily on partners with prior IOU or RTO modeling backgrounds. Generic SaaS-forecasting partners struggle to navigate the regulatory data structures and the specific load-forecasting horizons that utilities use.
Beyond data centers and the grid, Cheyenne carries a smaller but technically distinctive layer of ML demand. F.E. Warren Air Force Base, hosting the 90th Missile Wing and a substantial share of the US ICBM force, generates classified ML work that mostly runs through DoD prime contractors rather than open commercial procurement. Outside ML partners with security clearances and prior defense-contractor experience — particularly with Northrop Grumman on the Sentinel ICBM program, with Boeing on the legacy Minuteman III sustainment, or with the Air Force Global Strike Command at Barksdale — have credible angles for sub-tier engagements. State government ML, run through the Wyoming Department of Administration and Information and the Capitol-complex data infrastructure, covers Medicaid forecasting through the Wyoming Department of Health, tax-revenue modeling, and Department of Transportation forecasting along I-25, I-80, and the state highway network. The Wyoming Game and Fish Department occasionally funds wildlife-tracking ML on big-game herds across the state. Cheyenne LEADS, the county economic development organization, and Laramie County Community College's data analytics certificate program supply some of the local talent network. ML partners working in this Cheyenne tier need clearance experience for the F.E. Warren work and government-procurement experience for the state-agency work — neither transfers cleanly from a commercial SaaS background.
Mostly no. The hyperscale ML work runs centrally at Microsoft, Meta, and their core infrastructure-engineering teams, and procurement runs through corporate vendor management organizations, not through Cheyenne local IT. Outside ML partners rarely land core capacity-planning or PUE-optimization work directly at Microsoft North Bay or Meta Cheyenne. The realistic adjacent opportunities are in the colocation tier in the Cheyenne Logistics Park, with Black Hills Energy on the grid side, and with construction and facilities partners that support data-center buildouts. ML partners pitching the hyperscalers directly without an existing corporate relationship usually do not advance past the first conversation. The supporting ecosystem is more accessible and increasingly well-funded as the data-center concentration grows.
Several distinct pieces. Load-forecasting ML against hyperscale customers requires modeling that handles step-function load changes when new data-center buildings energize, AI-training-job batching patterns that drive intra-day load shape, and the unusual seasonal pattern of free-cooling availability in the High Plains climate. Demand-response ML covers behind-the-meter generation modeling for colocation operators with backup-generator participation in regional capacity markets. Transmission-side ML covers WAPA interconnection capacity and the western grid integration. Rate-design ML supports the utility's cost-of-service work for Wyoming Public Service Commission filings. ML partners need prior IOU, RTO, or transmission planner experience. Generic forecasting partners struggle with the regulatory data structures and the specific forecasting horizons utilities use for capacity planning.
Yes, but only through specific structures. Direct work with F.E. Warren or with Air Force Global Strike Command requires clearances and a defense-contractor relationship — typically as a subcontractor to Northrop Grumman on the Sentinel ICBM modernization program, to Boeing on the legacy Minuteman III sustainment, or to one of the Tier 1 IT integrators serving the base. ML partners with prior DoD work, particularly in nuclear-mission support, intelligence, surveillance, and reconnaissance, or in classified networks, have credible angles. The procurement timelines are long, the security overhead is substantial, and the realistic engagement cadence is multi-year rather than quarterly. Partners coming from purely commercial backgrounds rarely succeed in this tier without a clearance-bearing teaming partner.
The NCAR-Wyoming Supercomputing Center north of Cheyenne hosts the Cheyenne and now Derecho supercomputers and is one of the largest atmospheric and earth-science HPC facilities in the country. Most of the compute serves NCAR research and the broader university community, but the center's presence shapes the local ML talent ecosystem — researchers working on climate, weather, and atmospheric ML often have NCAR relationships and the experience of running on petascale HPC systems. ML partners with prior NCAR collaboration or related HPC backgrounds bring credibility to weather-forecasting, climate-modeling, and atmospheric-data ML engagements that smaller commercial firms struggle to offer. The University of Wyoming-NCAR partnership occasionally funds collaborative work that lands in Cheyenne or Laramie.
The University of Wyoming's School of Computing in Laramie supplies most of the senior ML and data-science talent that lands in Cheyenne, with the Advanced Research Computing Center providing applied HPC experience. Colorado State University in Fort Collins, an hour to the south on I-25, is another meaningful pipeline. Laramie County Community College feeds data-center technician and operations-analyst roles. The Denver metro is within reasonable commute or relocation distance for senior ML roles. Cheyenne does not have the local senior labor pool that Denver or Salt Lake City have, and engagements requiring senior modeling staff typically combine a Cheyenne or Laramie consultant with travel-in expertise from Denver, Boulder, or Fort Collins. ML partners need to be honest about which roles are staffable locally.
Get listed on LocalAISource starting at $49/mo.