enrichment coverage
277,904 39.2% extracted a value was found and written
149,677 21.1% no signal the pass ran and found nothing to extract
282,235 39.8% never enqueued the document was never handed to this pass
5 0 err in flight queued or running right now
why it skipped
no-entities-found 21,356 no-outcome-signal 18,027 no-case-no-and-no-defendant 16,216 no-court-match 15,710 no-sentence-detected 14,766 no-classified-amounts 14,010 no-statutes-found 13,590 no-disposition-detected 13,399 no-sanctions-detected 11,599 no-facts-extracted 6,506
how to get that number down — 142,093 field/document pairs still measurably missing, plus 4 fields not yet measurable
sentence-extract30,837(documents still missing this field) Enqueue the 18,658 documents this pass has never seen — coverage is bounded by the queue, not the extractor.disposition-extract30,837(documents still missing this field) Enqueue the 18,658 documents this pass has never seen — coverage is bounded by the queue, not the extractor.sanctions-extract20,184(documents still missing this field) Enqueue the 18,658 documents this pass has never seen — coverage is bounded by the queue, not the extractor.outcome-detect18,001(documents still missing this field) Regex found no signal. Route the skipped set through an LLM pass (pool-qwen-large) to recover the long tail.link-references15,916(documents still missing this field) Regex found no signal. Route the skipped set through an LLM pass (pool-qwen-large) to recover the long tail.statute-refs8,977(documents still missing this field) Regex found no signal. Route the skipped set through an LLM pass (pool-qwen-large) to recover the long tail.court-jurisdiction8,967(documents still missing this field) Regex found no signal. Route the skipped set through an LLM pass (pool-qwen-large) to recover the long tail.case-facts3,292(documents still missing this field) Enqueue the 26,194 documents this pass has never seen — coverage is bounded by the queue, not the extractor.
Skipped is not failed — the extractor ran and found no match. Two levers move it: enqueue the
documents a pass has never seen, and re-run the no-signal set through an LLM pass, which reads
prose a regex cannot. Documents whose body is thin or image-only need refetch-pdf / ocr-recover first, otherwise every downstream pass will
skip them again.
outcomes
settled 7,536 pleaded 3,207 charged 2,701 convicted 1,530 indicted 468 sentenced 331 dismissed 19 acquitted 8
courts
SDNY 10,192 CDCA 1,373 SDFL 1,337 DC 1,219 DMA 1,091 EDNY 835 NDCA 806 NDIL 729 DNJ 618 NDTX 586
entity kinds
person 56,145 company 32,078 organization 16,890 agency 9,586 scheme_term 5,547 location 1,829 court 976 unknown 394
coverage by workflow
| field | coverage | extracted | no signal | never run | queue | top skip reason |
|---|---|---|---|---|---|---|
| Statute refs | 20,21159.8% | 13,59040.2% | 00.0% | 0 | no-statutes-found 13,590 | |
| Court / jurisdiction | 18,09153.5% | 15,71046.5% | 00.0% | 0 | no-court-match 15,710 | |
| Keywords | 33,70199.7% | 1000.3% | 00.0% | 0 | no-keywords-extracted 100 | |
| Outcome | 15,77446.7% | 18,02753.3% | 00.0% | 0 | no-outcome-signal 18,027 | |
| Money classes | 19,79158.6% | 14,01041.4% | 00.0% | 0 | no-classified-amounts 14,010 | |
| Title formatted | 33,69899.7% | 00.0% | 1030.3% | 0 | — | |
| Title LLM | 10,70131.7% | 4121.2% | 22,68867.1% | 0 | llm-error:title-llm HTTP 503: {"error":{"type":"observability_failure","message":"gpumon-w 284 +3 | |
| Titles | 7752.3% | 1020.3% | 32,92497.4% | 0 | no title found via any tier (1-6) 102 | |
| Scheme | 32,32995.6% | 00.0% | 1,4724.4% | 0 | — | |
| Scheme detail | 30,85791.3% | 5971.8% | 2,3476.9% | 0 | detail-too-short 206 +3 | |
| Link references | 17,58552.0% | 16,21648.0% | 00.0% | 0 | no-case-no-and-no-defendant 16,216 | |
| Summary | 510.2% | 00.0% | 33,75099.8% | 0 | — | |
| Refetch PDF | 4061.2% | 5171.5% | 32,87897.3% | 0 | body-already-sufficient 462 +2 | |
| Entities | 11,12032.9% | 21,35663.2% | 1,3253.9% | 0 | no-entities-found 21,356 | |
| Mentions | 11,15733.0% | 2,0736.1% | 20,57160.9% | 0 | no-entities-in-triples 2,073 | |
| OCR recovery | 340.1% | 4631.4% | 33,30498.5% | 0 | body-already-sufficient 463 | |
| Sentences | 3471.0% | 14,79643.8% | 18,65855.2% | 0 | no-sentence-detected 14,766 +1 | |
| Dispositions | 1,7145.1% | 13,42939.7% | 18,65855.2% | 0 | no-disposition-detected 13,399 +1 | |
| Sanctions | 3,51410.4% | 11,62934.4% | 18,65855.2% | 0 | no-sanctions-detected 11,599 +1 | |
| Case facts | 1,0963.2% | 6,50619.2% | 26,19477.5% | 5 | no-facts-extracted 6,506 | |
| Case facts LLM | 14,95244.2% | 1440.4% | 18,70555.3% | 0 | no-llm-facts 102 +3 |
extracted ran, no signal never enqueued
mentions: 229,142 docs with money classified: 33,801 settlement Σ: $314749.7M disgorgement Σ: $305333.2M civil penalty Σ: $2444553.8M victim loss Σ: $23517940.1M total victims: 40,834,048
documents per month
SEC LitSEC PRDOJ SDNY
documents per year
SEC LitSEC PRDOJ SDNY
kinds
top subjects
top objects
top verbs
sanctions & losses by magnitude fetching· 0 B
no labelled amounts detected
sanction & loss size per year fetching· 0 B
no labelled amounts by year
case filings
filings by source & type
filings per year by type
complaintjudgmentinformationindictmentsuperseding indictment
first overt act → indictment lag
Backed by triples extraction once the processor parses overt-act dates from complaint
narratives and aligns them against indictment published_at. Replace with
real data when case_facts.first_act_date populates.
defendants per case
case dispositions fetching· 0 B
sentence length distribution fetching· 0 B
corporate fraud focus
fraud type distributionsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
fraud types per yearsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
Pump-and-dumpInsider tradingUnregisteredMkt manipulationAccounting fraud
fraud type by sourcesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
top defendants — corporatesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
top target companiessynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
top tickerssynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
criminal analysis — offence, instrument, role
alleged conductsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
investment vehiclessynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
positions of trustsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
securities fraud — detail view
securities money bucketssynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
securities statutessynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
pump-and-dump trendsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
unregistered securities trendsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
insider trading trendsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
market manipulation trendsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
microcap + shell-company schemes
microcap flag ratesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
34.0%
shell flag ratesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
21.0%
microcap/shell overlapsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
12.0%
microcap by fraud sub-typesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
shell-co by fraud sub-typesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
microcap/shell promoterssynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
federal crime distribution
federal crime distributionsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
crime bucket trendsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
Securities fraudWire / mail fraudMoney launderingFCPA
case-no cardinalitysynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
14,320
caption parse coveragesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
87.0%
litigators
top AUSAssynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
top SEC counselsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
top judgessynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
litigator fraud-type mixsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
civil penalty distributionsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
disgorgement distributionsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
signal & anomaly
benford leading digitsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
benford deviation by yearsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
summary quality pulsesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
82.0%
OCR quality pulsesynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
74.0%
body length distributionsynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
filing calendar
filings per daysynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
sec-pr
sec-litreleases
doj-sdny
scheme calendar
schemes per daysynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
microcap-fraud
accounting-fraud
pump-and-dump
insider-trading
ponzi
entity co-occurrence graph
entity network
loading…
scheme → outcome flow
scheme → outcome
loading…
scheme dollar packing
scheme & amount packing
loading…
cross-corpus: who is charged, and with whom
cross-corpus networksynthetic Warning: this panel shows synthetic placeholder data, not real enforcement figures.
Loading cross-corpus data...
top 10 classified schemes
all classified schemes
coverage summary
corpus shape & case outcomes
case dispositionsfetching· 0 B
sources
SEC Lit SEC PR DOJ SDNY