The State Calls It "Staff Aug." We Mapped It Against H-1B Filings.

Methodology

Four steps, each checked against the others before anything here was treated as solid:

1. Broaden past the word "ITSAC." Texas's Legislative Budget Board Contract Management System (cms.lbb.texas.gov) lets agencies write their own free-text contract subject line, and most don't write "ITSAC" even when the underlying program is the state's DIR-administered IT Staff Augmentation Contract. A sample pull confirmed agencies label the identical program "Staff Aug," "IT Staff Aug Contract," "FY26AUGR Pos#...," "IT Services-Staff Augmentation," "Temporary Personnel Services," and dozens of other variants. We matched on Subject REGEXP 'staff.{0,3}aug|itsac|augr[^a-z]|it staff' combined with NIGP commodity code 962-69 ("Personnel Services, Temporary") as a structured backstop. That took the active-vendor universe from 167 (literal "ITSAC" match only) to 414, and the agency count from roughly 20 to 62.

2. Match vendor names to H-1B sponsors. State contract vendor names and DOL's H-1B employer names are typed independently by different people at different times, so we normalized both (uppercase, corporate-suffix stripped, punctuation removed) and matched as a prefix rather than requiring an exact string match. That prefix approach is deliberately permissive, so we added a second filter: reject any match where the leftover text after the vendor's core name isn't a real corporate-entity word (LLC, Inc., Corp., etc.). That filter caught, and excluded, real false positives — "MEDCOMPS CORPORATION" is not the same company as ITSAC vendor "medcomp Corporation" (the "S" is a different business), and "Apex System Solutions" is not "Apex Systems."

3. Aggregate by vendor, not by name spelling. Some vendors file H-1B petitions under more than one name variant across different case years. Grouping by raw employer-name string (our first pass) made "Resource Integrators LLC" look Texas-concentrated in one filing and nationally spread in another — combined, it's neither exclusively. We re-aggregated at the vendor-entity level before drawing any conclusion about a vendor's geographic footprint.

4. Verify addresses, don't assume them. For vendors whose full H-1B footprint is 100% Texas, we pulled individual, case-level LCA records — job title, filed worksite street address, wage, certification status, decision date — and checked the more distinctive addresses against public sources (the Texas Comptroller's own agency-address directory, and agency contact pages), rather than relying on general familiarity with Austin geography. That check mattered: our first guess about one address (assuming it was the Comptroller's office) was wrong. The verified source is cited at each address below.

The Universe: 414 Vendors, 62 Agencies, One Program With Many Names

Broadening the search past the literal word "ITSAC" more than doubled the visible vendor pool and added agencies that were essentially invisible to a keyword-only search — most notably the Health and Human Services Commission and the Employees Retirement System, both heavy users of the program under non-"ITSAC" subject lines:

Vendor and agency counts reflect Active-status contracts in Texas's LBB Contract Management System as pulled September 3, 2026. "Vendors also H-1B sponsors" counts distinct normalized vendor entities with at least one FY2025-FY2026 certified LCA petition, after excluding name-collision false positives.

Nine Vendors Whose Entire H-1B Program Is Texas

Most of the 122 matched vendors are large, multi-state staffing and consulting firms — their H-1B sponsorship reflects a national client book, and overlap with a Texas state contract proves very little about any specific worker's assignment. But nine vendors' entire FY2025-FY2026 H-1B sponsorship — every certified petition, every filed worksite — is in Texas and nowhere else. For a company whose whole U.S. visa-sponsorship program is Texas-only, the odds that those sponsored workers are the same people fulfilling that company's Texas state contracts are much higher than for a national firm with thousands of unrelated clients.

Two Building Matches, Verified — One Open Lead

For each Texas-only vendor, we pulled the individual street addresses their certified LCA petitions list as the worker's actual worksite. Many read as residential subdivision streets — a strong signal of remote/telework arrangements, not on-site placement, and we treat those as inconclusive. But a recurring handful are recognizable government-office-style addresses. We checked the most distinctive of them against the Texas Comptroller's own agency-address directory and agency contact pages before drawing any conclusion:

The Agency's Own Paperwork Names the Worker

One byproduct of broadening the search past "ITSAC": the Health and Human Services Commission's own purchase-order subject lines routinely embed the placed contractor's name and an internal position number, functioning as an informal staff roster inside otherwise-public procurement data. A sample, all Active-status, all tied to vendors in this report:

Full names are truncated in the state's own source data (the subject-line field is length-limited). We have not attempted to expand these to full legal names or cross-match them to a specific LCA case — DOL's public H-1B disclosure data does not include worker names, so there is no way to directly confirm whether any specific named contractor here is the same individual as any specific H-1B filing above. We list these because the state's own procurement data already functions as a partial staffing roster, independent of anything in this report's H-1B analysis.

What Roles Are Actually Being Filed For

Every certified LCA in this dataset carries a job title. Pooled across all 122 matched vendors, 27,279 petitions span 2,308 distinct titles — but that pool is dominated by two of the largest national consulting firms in the match set. Deloitte Consulting alone accounts for 10,713 of the 27,279 petitions (39%), and its internal job-ladder titles — "Senior Consultant," "Specialist Master," "Solution Specialist," generic "Manager" — read like corporate rank names, not descriptions of any specific Texas placement. CGI Technologies contributes another 1,973. Treat the aggregate below as describing the vendor pool's overall workforce, not the Texas state contracts specifically:

By contrast, the individual case-level filings we could tie to a specific verified Texas placement earlier in this report — Business Analyst, Software Engineer, Data Scientist, PeopleSoft Systems Administrator, Data Analyst/Report Writer, Tester — are ordinary IT staff-aug job titles, not consulting-firm rank names. The gap between those two lists is itself the clearest illustration of why vendor-level aggregates need to be read differently for a handful of specialized shops than for national consultancies with thousands of unrelated clients.

Investigative Assessment

What This Report Does Not Show

💬 Comments