How chidokato5376/finguard-ai-platform works
FinGuard AI is a transaction-risk management platform: prepared transaction data feeds risk scoring, suspicious cases are explained with behavioral and network context, and QAPE prioritizes alerts for human review. Its main project adds DRRE training and quantum-method research; the public demo instead uses heuristic scoring. The dashboard is the user-facing entry point, with story logic coordinating data, scoring and prioritization services. Edges follow sampled code, SOURCE INDEX references or README statements; unsampled page and quantum internals are represented only where documented.
Main entry points: story.py Story layer · train.py DRRE training · app.py Dashboard app · scoring_service.py Scoring service
Components 22
Dashboard
- Dashboard app
Finguard/src/dashboard/app.py - Dashboard pages
Finguard/src/dashboard/app_pages - Story layer
Finguard/src/dashboard/story.py
Transaction Data
- Data service
Finguard/src/dashboard/data_service.py - Data loaders
Finguard/src/data/loaders.py - Preprocessing
Finguard/src/data/preprocessing.py - Data schema
Finguard/src/data/schema.py - Synthetic data
Finguard/src/data/synthetic_generator.py
Risk Intelligence
- Scoring service
Finguard/src/dashboard/scoring_service.py - DRRE model
Finguard/src/drre/model.py - DRRE components
Finguard/src/drre - Risk scoring engine
Finguard/src/scoring/risk_scoring.py - Explainability
Finguard/src/explainability/explain.py - Graph utilities
Finguard/src/dashboard/graph_utils.py
Alert Prioritization
- QAPE service
Finguard/src/dashboard/qape_service.py - Alert QUBO
Finguard/src/qape/qubo.py - QAPE solvers
Finguard/src/qape/solvers.py - Quantum methods
Finguard/src/quantum
Model Research
- DRRE training
Finguard/src/drre/train.py - Training losses
Finguard/src/drre/losses.py - Evaluation metrics
Finguard/src/evaluation/metrics.py
Other components
- Bank analyst
Connections 24
- Bank analyst → Dashboard app: uses
README.md - Dashboard app → Data service: lists sources
Finguard/src/dashboard/app.py - Dashboard app → Story layer: loads results
Finguard/src/dashboard/app.py - Dashboard app → Dashboard pages: routes to
Finguard/src/dashboard/app.py - Dashboard pages → Story layer: presents
Finguard/src/dashboard/app_pages/uu_tien.py - Story layer → Data service: reads data
Finguard/src/dashboard/story.py - Story layer → Scoring service: gets scores
Finguard/src/dashboard/story.py - Story layer → QAPE service: gets priorities
Finguard/src/dashboard/story.py - Data service → Data loaders: reads processed data
- Preprocessing → Data loaders: loads sources
- Synthetic data → Data schema: uses schema
Finguard/src/data/synthetic_generator.py - Scoring service → DRRE model: runs model
Finguard/src/dashboard/scoring_service.py - Scoring service → Risk scoring engine: scores embeddings
Finguard/src/dashboard/scoring_service.py - Scoring service → Graph utilities: builds graph
Finguard/src/dashboard/scoring_service.py - DRRE model → DRRE components: uses components
Finguard/src/drre/model.py - Explainability → DRRE model: interprets signals
- QAPE service → Alert QUBO: builds alerts
Finguard/src/dashboard/qape_service.py - QAPE service → QAPE solvers: solves batch
Finguard/src/dashboard/qape_service.py - QAPE solvers → Alert QUBO: converts problem
Finguard/src/qape/solvers.py - QAPE solvers → Quantum methods: includes QAOA
README.md - DRRE training → DRRE model: trains model
Finguard/src/drre/train.py - DRRE training → Training losses: computes losses
Finguard/src/drre/train.py - DRRE training → Evaluation metrics: evaluates model
Finguard/src/drre/train.py - DRRE training → Risk scoring engine: uses engine
Finguard/src/drre/train.py