Build a Transfer Network and Legitimate-Activity Context Desk

A financial crime analyst reviewing linked transfer alerts.

Investigate transfer clusters while preserving legitimate payroll, household, and merchant explanations. Show why a shared connection alone does not establish a money-mule network.

Key: transaction_id with account_id graph nodes.

Fields: sender, recipient, amount, event_time, relationship_evidence, provenance.

Common infrastructure links differ from ownership links. Reversed transfers retain both original and reversal IDs.

Build time-bounded graphs with duplicate-edge controls.

Trace candidate transfer paths and alternative business explanations.

Record evidence sufficient for review and evidence still missing.

Export bounded subgraphs with investigator annotations.

A time-filtered graph with path evidence, legitimate-context panels, and an analyst hypothesis ledger.

A reproducible graph extraction pipeline.

Case packets that preserve alternative explanations and unknown intent.

Runnable local application, setup commands, and a dependency lockfile.

Versioned input data, data dictionary, acquisition instructions, and rights notes.

Calculation traces, source references, assumption records, and exported review artifacts.

A deterministic baseline and tests for normal, incomplete, contradictory, and failure-recovery cases.

A recorded complete workflow with a reproducible manifest and explicit limitations.

Graph queries cannot use edges after the decision cutoff.

Repeated ingestion does not inflate transfer amounts.

Shared infrastructure alone never becomes a confirmed ownership relation.

Classify review needs separately from intentional participation or criminal guilt.

Two clusters share the same transfer total, but one has duplicate ingestion events. Deduplication must change its apparent volume.

The attached starter files are prototype inputs. Expected starter answers are visible. They are not independent holdouts.

Full target: Use a permitted PaySim extract for simulated transfer structure. Author separate, labeled fictional relationship and business-context records rather than claiming they exist in PaySim.

Complete target contracts and expert review remain pending. Public observations are not historical release vintages.

NetworkX, DuckDB, Plotly

Use an available model appropriate to the task. Record provider, model identifier, version or retrieval date, settings, cost, and latency. Use deterministic code for calculations. Models may extract, retrieve, or explain evidence. They must not invent missing records, override approval rules, or claim professional authority.

Synthetic graph patterns are investigative leads, not evidence of intent or a real money-laundering finding.

No fabricated records, source citations, professional endorsements, or expert-adjudication claims.

No future-data leakage into dated decisions, model fitting, or historical valuation inputs.

No unauthorized payments, account restrictions, regulatory filings, or production changes.

No automated appraisal certification, tax-compliance determination, or unsupported fraud accusation.

No concealed uncertainty or silently invented inputs required for a numerical conclusion.

Correct domain calculations and justified methods with documented assumptions. Weight: 30.

Source lineage, temporal integrity, rights checks, and explicit synthetic-data labels. Weight: 25.

A working application with review, export, error handling, and reproducible state. Weight: 20.

Adversarial cases, uncertainty, missing-data behavior, and sensitivity analysis. Weight: 15.

Evidence-linked conclusions that distinguish facts, assumptions, and unresolved judgments. Weight: 10. Score each dimension from zero to two. Weighted score = sum(weight * score / 2). A failed safety gate blocks acceptance.