Running pattern detector against the corpus
The dataset shows entities that received substantial federal funding then ceased appearing in the corpus — flagging recipients that went silent after the money flowed.
Aggregated from the matches below. Severity bands are calibrated per pattern; counts and dollar totals reflect this snapshot only.
Glassbox surfaces correlations and observations, not findings of misconduct. Severity is a calibrated read on how strongly the dataset matches the pattern definition — a starting point for review, not a verdict.
The match meets the pattern threshold but the signal is the lower band. Normal review queue.
The match exceeds the routine threshold. Worth a closer look — pull the source rows, walk the citation.
The match is in the highest band of the calibrated scale. Recommend: prioritise for active review.
Severity is computed from the detection signal · total funding ≥ $500K · last agreement_start_date < CURRENT_DATE − 36 months · is_amendment = false
The dataset shows First Nations Health Authority received $8,219,808,503 across 4 federal agreements from 3 departments, with the most recent agreement dated 2023-04-01 — 3.1 years of subsequent silence in the corpus.
The dataset shows Batch report│Rapport en lots received $966,961,115 across 29 federal agreements from 1 departments, with the most recent agreement dated 2023-03-31 — 3.1 years of subsequent silence in the corpus.
The dataset shows UNICEF - United Nations Children's Fund received $908,343,865 across 69 federal agreements from 1 departments, with the most recent agreement dated 2023-03-27 — 3.1 years of subsequent silence in the corpus.
The dataset shows 1000511515 Ontario Inc. (BN 000000000) received $700,000,000 across 1 federal agreements from 1 departments, with the most recent agreement dated 2023-04-01 — 3.1 years of subsequent silence in the corpus.
The dataset shows Employment and Social Development Canada (ESDC) received $638,259,730 across 3 federal agreements from 1 departments, with the most recent agreement dated 2023-04-01 — 3.1 years of subsequent silence in the corpus.
The dataset shows ENBRIDGE GAS INC received $613,437,876 across 1 federal agreements from 1 departments, with the most recent agreement dated 2022-08-26 — 3.7 years of subsequent silence in the corpus.
The dataset shows ONTARIO MINISTRY OF THE ATTORNEY GENERAL | MINISTÈRE DU PROCUREUR GÉNÉRAL DE L'ONTARIO received $572,615,417 across 7 federal agreements from 1 departments, with the most recent agreement dated 2023-04-01 — 3.1 years of subsequent silence in the corpus.
The dataset shows Umicore Canada Inc. (BN 105447841) received $551,349,400 across 1 federal agreements from 1 departments, with the most recent agreement dated 2022-07-04 — 3.8 years of subsequent silence in the corpus.
The dataset shows GOVERNMENT OF QUEBEC received $543,579,000 across 2 federal agreements from 1 departments, with the most recent agreement dated 2022-08-26 — 3.7 years of subsequent silence in the corpus.
The dataset shows NextStar Energy Inc. (BN 730481009) received $500,000,000 across 1 federal agreements from 1 departments, with the most recent agreement dated 2022-09-30 — 3.6 years of subsequent silence in the corpus.
Once a match is identified, the dataset suggests these review actions. Glassbox does not prescribe outcomes — these are calibrated options for funders, auditors, and program officers, framed as correlations to investigate.
Cross-reference the recipient legal name with the relevant federal/provincial corporate registry. An active corporate status with no recent grant activity may indicate program completion (normal); a dissolved or in-default status with prior public funding warrants review.
If the recipient is a registered charity (CRA-registered), pull the most recent T3010 filing date. A gap exceeding 18 months is a regulatory issue independent of the federal-funding silence.
Did the funded program complete? Was the final report submitted? Were funds returned, if applicable? Silence in the corpus is not evidence of incompletion, but it is the signal to verify.
Some zombies are just program endings; others are entity disappearances after public money flowed. Close the match with reasoning, or move to active monitoring if outcomes are unclear.
These are calibrated suggestions, not directives. The dataset shows the pattern; the funder decides the response.