Skip to main content
Challenge

Build an Investigation Case Dossier Compiler with Evidence Provenance

Compile structured investigative case files linking transactions, legal entity records, and open-source intelligence. Track data provenance timestamps and explicit evidentiary gaps to prepare defensible case summaries.

CybersecurityHosted by Vera
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Compile structured investigative case files linking transactions, legal entity records, and open-source intelligence. Track data provenance timestamps and explicit evidentiary gaps to prepare defensible case summaries.

Senior AML investigator

Assemble multiple evidence artifacts including normalized invoice rows, GLEIF entity extracts, and transaction hops.

Bind cryptographic hash signatures and source timestamps to every evidence item to preserve analytical provenance.

Identify disconnected hypothesis links where narrative claims lack supporting transaction artifacts.

Generate a finalized case preparation bundle with visual provenance graphs and missing evidence disclosures.

Evidence binder workspace showing linked supporting documents, timestamped provenance breadcrumbs, and gap warnings.

Create 20 mock investigative cases containing heterogeneous evidence objects (PDF text extracts, JSON transaction records, LEI profiles). Starter examples below are synthetic. They are not downloaded records.

GLEIF API Integrates GLEIF reference records and parent-child relationship models into structured case evidence collections.

pydantic-ai, jinja2, cryptography. Pydantic AI enforces rigorous provenance metadata schema validation across diverse evidence types. Local or hosted LLM API used strictly to summarize linked evidence into factual briefing memos.

Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

Generates SHA-256 hash digests for all incorporated evidence artifacts.

Automatically flags any narrative assertion that cannot be mapped to a recorded artifact ID.

Exports a consolidated JSON case dossier compliant with standardized internal audit formats.

```json [ { "input": { "case_id": "CASE-101", "artifacts": [ { "artifact_id": "ART-1", "type": "GLEIF_RECORD", "content": "LEI: 5493006MHB84DD0ZWV18, Apex Global Logistics Inc" } ], "narrative_assertions": [ { "claim": "Apex Global Logistics Inc is a registered legal entity.", "linked_artifact_id": "ART-1" } ] }, "expected_output": { "provenance_verified": true, "unsupported_claims": [], "status": "READY_FOR_REVIEW" } }, { "input": { "case_id": "CASE-102", "artifacts": [ { "artifact_id": "ART-1", "type": "TRANSACTION_ROW", "content": "Transfer 50,000 USD from A to B" } ], "narrative_assertions": [ { "claim": "Party B forwarded funds to Party C in offshore jurisdiction.", "linked_artifact_id": null } ] }, "expected_output": { "provenance_verified": false, "unsupported_claims": [ "Party B forwarded funds to Party C in offshore jurisdiction." ], "status": "EVIDENCE_GAPS_DETECTED" } } ] ```

Assists analytical preparation of internal dossiers; forbids automated dispatch to external law enforcement or regulatory authorities. Use mock external writes. Require human review before any real deployment.

Rules

Participation guide

Assists analytical preparation of internal dossiers; forbids automated dispatch to external law enforcement or regulatory authorities. Use mock external writes and human review.

Delivery guide

How work is evaluated

Evaluation

Generates SHA-256 hash digests for all incorporated evidence artifacts. Automatically flags any narrative assertion that cannot be mapped to a recorded artifact ID. Exports a consolidated JSON case dossier compliant with standardized internal audit formats.

Submission format

Runnable source, pinned dependencies, setup commands, working interface, example outputs, recorded demo, tests, and action logs.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
4 scoring checks
Binary
1 pass or fail dimensions
Ordinal
3 scaled dimensions
Dimension 1

Complete workflow

Assemble multiple evidence artifacts including normalized invoice rows, GLEIF entity extracts, and transaction hops. Bind cryptographic hash signatures and source timestamps to every evidence item to preserve analytical provenance. Identify disconnected hypothesis links where narrative claims lack supporting transaction artifacts. Generate a finalized case preparation bundle with visual provenance graphs and missing evidence disclosures.

Score scale
012
Score 0

Missing or incorrect.

Score 1

Partly complete, with documented gaps.

Score 2

Complete, with reproducible evidence.

Dimension 2

Correctness and evidence

Generates SHA-256 hash digests for all incorporated evidence artifacts. Automatically flags any narrative assertion that cannot be mapped to a recorded artifact ID. Exports a consolidated JSON case dossier compliant with standardized internal audit formats.

Score scale
012
Score 0

Missing or incorrect.

Score 1

Partly complete, with documented gaps.

Score 2

Complete, with reproducible evidence.

Dimension 3

Review interface

Evidence binder workspace showing linked supporting documents, timestamped provenance breadcrumbs, and gap warnings.

Score scale
012
Score 0

Missing or incorrect.

Score 1

Partly complete, with documented gaps.

Score 2

Complete, with reproducible evidence.

Dimension 4

Safe execution

Assists analytical preparation of internal dossiers; forbids automated dispatch to external law enforcement or regulatory authorities.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Assemble multiple evidence artifacts including normalized invoice rows, GLEIF entity extracts, and transaction hops.

  • Bind cryptographic hash signatures and source timestamps to every evidence item to preserve analytical provenance.

  • Identify disconnected hypothesis links where narrative claims lack supporting transaction artifacts.

  • Generate a finalized case preparation bundle with visual provenance graphs and missing evidence disclosures.

Resources and assets

Reference links and supporting material

Dataset notes

Create 20 mock investigative cases containing heterogeneous evidence objects (PDF text extracts, JSON transaction records, LEI profiles). Supplied starter examples are synthetic, not source extracts.

Challenge setup

Configuration summary for this challenge.

Challenge input

Input details are not configured.

Pydantic AI

Typed Python agent framework.

Evaluated output

No output checks are configured.

Evaluation checks
  • Assemble multiple evidence artifacts including normalized invoice rows, GLEIF entity extracts, and transact...
  • Generates SHA-256 hash digests for all incorporated evidence artifacts. Automatically flags any narrative a...
  • Evidence binder workspace showing linked supporting documents, timestamped provenance breadcrumbs, and gap...
Configured evaluation evidence
  • 2 public reference cases
Configured runtime
  • 2 configured tools

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • Pydantic AI · Optional
Orchestration
  • Pydantic AI · Optional

Evaluation contract

  • Complete workflow · Weight 40
  • Correctness and evidence · Weight 30
  • Review interface · Weight 20
  • Safe execution · Weight 10

Reference cases

  • 2 public reference cases

Workflow tasks

No workflow tasks are configured.

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build an Investigation Case Dossier Compiler with Evidence Provenance