Glassbox VA · pattern catalog

Follow the money.

13 named patterns across U.S. Department of Veterans Affairs contracts and grants. Each surfaces specific records that match. Each output cites every source row.

5 veteran-business & eligibility checks · 8 spending patterns

SDVOSB Set-Aside to an Uncertified Firm

Live

The dataset shows contracts awarded under a service-disabled or veteran-owned set-aside or sole-source authority to a recipient whose UEI has no matching SBA veteran certification active on the award date.

Signal · app.award_current contracts with set_aside_code ∈ {SDVOSBC, SDVOSBS, VSA, VSS} (and SDVOSB/VOSB set-aside text) · first_action_date ≥ 2023-01-01 (SBA VetCert era) · obligated_total ≥ $25K · no sba.cert row for recipient_uei with cert_type matching the set-aside (SDVOSB for SDVOSB set-asides; SDVOSB or VOSB for veteran set-asides; JV variants accepted) and cert_start ≤ first_action_date ≤ COALESCE(cert_exit, ∞) · signal = net obligations $ · weaker when a certification exists at another date

1,582 records match

Veteran-owned business & eligibility · 38 U.S.C. 8127 (Veterans First) · 13 CFR 128 (VetCert) · GAO High Risk List: VA Acquisition Management

SDVOSB Flag Without Certification

Live

The dataset shows contracts whose federal record flags the recipient as service-disabled-veteran-owned while the recipient's UEI has no SBA SDVOSB certification on or before the award date.

Signal · app.award_current contracts with sdvosb_flag = true · first_action_date ≥ 2023-01-01 · obligated_total ≥ $25K · no sba.cert SDVOSB / SDVOSB JV row for recipient_uei with cert_start ≤ first_action_date · signal = net obligations $ · set-aside awards are reported by the set-aside detector instead

1,896 records match

Veteran-owned business & eligibility · 13 CFR 128 (VetCert) · FAR 4.1202 representations · VA OIG veteran-owned small business reviews

Award to an Excluded Party

Live

The dataset shows awards whose first action falls inside a SAM.gov exclusion period for the same recipient — matched by UEI, or by normalised name and state at lower strength.

Signal · sam.exclusion active_date ≤ first_action_date ≤ COALESCE(termination_date, ∞) · match on exclusion.uei = recipient_uei (strong) or exclusion.name_norm = normalised recipient_name AND exclusion.state = recipient_state (weak, penalised) · firm classifications only · obligated_total ≥ $10K · signal = net obligations $

1 records match

Veteran-owned business & eligibility · FAR 9.405 (effect of listing) · 2 CFR 180 · VA OIG suspension & debarment reviews

Certified-Firm Address & Contact Clusters

Live

The dataset shows groups of SBA-certified veteran-owned firms that share a street address, phone number or principal and together hold VA awards — a pattern associated with pass-through and rent-a-vet schemes that the certifying agency can verify.

Signal · sba.cert active VOSB/SDVOSB firms grouped by (address_norm, zip5), phone_norm, or principal name · ≥ 3 distinct UEIs per group · known registered-agent / virtual-office addresses (sba.address_denylist, e.g. 30 N GOULD ST 82801) and any key shared by > 25 registrants excluded · group VA obligations ≥ $1M · signal = distinct firms in the group

56 records match

Veteran-owned business & eligibility · 13 CFR 128.203 (control by the veteran) · 38 U.S.C. 8127 · VA OIG SDVOSB eligibility findings

Registration Lapse After Award

Beta

The dataset shows recipients whose SAM.gov registration expired or disappeared from the public extract within a year of a large VA award.

Signal · recipient VA obligations ≥ $1M · sam.entity.registration_expiration_date < last public extract date, or last_seen_src_date < latest extract · lapse within 12 months after the recipient's latest award action · signal = months between last award and lapse

93 records match

Veteran-owned business & eligibility · FAR 4.1102 (SAM registration) · GAO High Risk List: VA Acquisition Management

Sole-Source & Modification Creep

Live

The dataset shows contracts that were not competed and whose net obligations grew at least threefold beyond the first transaction through modifications.

Signal · app.award_current contracts · base_obligated (first transaction's federal_action_obligation) ≥ $100K · obligated_total (SUM of per-modification deltas) / base_obligated ≥ 3.0 AND ≥ p90 of growth ratios among modified contracts with the same PSC · not competed: extent_competed_code ∈ {B not available, C not competed, G not competed under SAP, NDO non-competitive delivery order} or, when no competition code is recorded, an other_than_full_and_open authority · ≥ 1 modification · signal = growth ratio

1,668 records match

Spending pattern · FAR 6.3 (other than full and open competition) · GAO High Risk List: VA Acquisition Management

Vendor Concentration

Live

The dataset shows contracting-office × product-or-service cells where a single vendor or small group receives a disproportionate share of obligations — incumbency replacing competition.

Signal · per (awarding office, PSC) over the rolling 36 months to the corpus as-of date · cell obligations ≥ $10M · ≥ 3 recipients · HHI = Σ recipient share² × 100² ≥ p90 of cell HHIs within the administration AND ≥ 1500 floor · bands ≥5000 extreme, ≥2500 highly, ≥1500 moderately concentrated · signal = HHI

60 records match

Spending pattern · FAR 6.1 (full and open competition) · GAO High Risk List: VA Acquisition Management

Zombie Recipients

Live

The dataset shows recipients that received substantial VA obligations and then stopped appearing in VA award data for three years or more.

Signal · recipient net obligations ≥ p90 of recipient totals within its primary administration AND ≥ $500K floor · latest action_date < as-of − 36 months, as-of = latest award action ≤ today · awards to individuals excluded · signal = years silent

3,589 records match

Spending pattern · Payment Integrity Information Act of 2019 · OMB Circular A-123 Appendix C

Ghost Capacity

Live

The dataset shows recipients of substantial VA obligations that cannot be tied to a SAM.gov registration — no valid UEI on the award, or a UEI absent from the SAM.gov entity extract.

Signal · app.award_current, awards to individuals excluded · recipient_uei IS NULL or is_placeholder_uei, or recipient_uei not in sam.entity (checked only when sam.entity is loaded) · recipient net obligations ≥ p90 of recipient totals within the administration AND ≥ $500K · signal = net obligations $

478 records match

Spending pattern · FAR 4.1102 (SAM registration) · Payment Integrity Information Act of 2019

Modification Purpose Drift

Live

The dataset shows awards whose current description shares few keywords with the original — the award may have drifted from its initial purpose.

Signal · Jaccard token similarity (description_initial, description_current) < 0.30 after dropping boilerplate modification tokens (option, admin, mod, exercise, funding, incremental …) · ≥ 3 transactions · net obligations ≥ $250K · signal = 1 − similarity

13,238 records match

Spending pattern · FAR 43.1 (contract modifications, scope) · GAO High Risk List: VA Acquisition Management

Duplicative Funding

Live

The dataset shows recipients holding both VA contracts and VA assistance awards for overlapping purposes — the same UEI paid through two instruments where the funder can check for duplication.

Signal · same recipient_uei with ≥ $100K in VA contracts AND ≥ $100K in VA assistance · overlap measured on shared description keywords between the two instruments' awards (Jaccard ≥ 0.2) or a known purpose pairing (homeless services PSC/NAICS ↔ 64.024 GPD / 64.033 SSVF) · signal = smaller of the two totals $

181 records match

Spending pattern · GAO duplication, overlap & fragmentation reports · 31 U.S.C. 712 note

Related Parties

Live

The dataset shows VA recipients that share a registered address, phone number or principal with other VA recipients across SAM.gov and SBA records — distinct registrations that may be affiliated.

Signal · link recipients by sam.entity (address_norm, zip5), sba.cert phone_norm, or sba.cert principal name · registered-agent addresses (sba.address_denylist) and keys shared by > 25 registrants (multi-tenant buildings) excluded · ≥ 3 distinct UEIs, each with VA obligations · group obligations ≥ $1M · signal = distinct linked recipients

233 records match

Spending pattern · 13 CFR 121.103 (affiliation) · FAR 9.104-3 · VA OIG procurement integrity reviews

Geographic Reconciliation Variance

Live

VA's Geographic Distribution of VA Expenditures (GDX) and USAspending.gov show different amounts for the same state, fiscal year and spending category, by more than the tolerance band set by all states that year. GDX reports expenditures; USAspending reports obligations.

Signal · gdx.reconciliation (scripts/va/gdx.ts): state × fiscal year × category for Compensation & Pension, E&VRE, Insurance and Construction only (Medical Care, GOE, Loan Guaranty and Memorial Benefits are not comparable) · ratio = USAspending obligations of the crosswalked assistance listings / PSC Y contracts ÷ GDX expenditure · band = Tukey fences exp(Q1 − 1.5·IQR, Q3 + 1.5·IQR) over ln(ratio) of all compared states that year (≥ 10 states) · variance when outside the band AND |ratio − median| ≥ 5 points AND |gap| ≥ $1M · GDX < $1M not compared · years compared only when all 12 months of USAspending are loaded · signal = |gap| $

53 records match

Spending pattern · VA NCVAS Geographic Distribution of VA Expenditures (GDX) · DATA Act (Pub. L. 113-101) award reporting · OMB Circular A-11 (obligations vs outlays)

Match counts · totals from the pattern store (see each pattern page footer for live-table vs snapshot)

Where the patterns come from

Each pattern responds to a published oversight concern: GAO's High Risk List entries for VA acquisition management and VA health care, VA Office of Inspector General findings on veteran-owned small business eligibility, and the payment-integrity framework of the Payment Integrity Information Act and OMB Circular A-123 Appendix C. Thresholds, sources and known false positives are listed on every pattern page.

Data sources & lineage →

Calibrated language

Glassbox VA surfaces correlations and observations, not causal claims. Every match carries an audit token, an evidence array citing source rows, and language reviewed against the calibration sweep (no “fraud”, no “clearly shows”, no “should have”).

Read the methodology →