CypherRing is a high-performance forensics engine engineered to expose money muling networks. The application transforms complex transaction data into graph structures to autonomously detect illicit financial patterns.
- Vaibhav Singh (Team Lead) — Lead Architect & Graph Logic 🏗️
- Antra Priyadarshini — UX/UI & Interactive Graph Visualization 🎨
- Srishti Maurya — Algorithm Optimization & Technical Documentation 📝
Our engine specifically targets the mandatory forensic patterns required by RIFT 2026:
- Logic: Identification of closed-loop chains (3 to 5 hops) where funds eventually return to the source account to obscure the trail.
- Fan-in: Detection of a single aggregator account receiving funds from 10+ distinct senders.
- Fan-out: Identification of a single sender dispersing funds to 10+ distinct receivers.
- Temporal Analysis: High-priority flagging for suspicious transactions occurring within a critical 72-hour window.
- Logic: Exposing intermediate "shell" accounts characterized by low transaction counts (2-3 total) that serve as transit points for layering.
- Backend: Python (Flask)
- Frontend: HTML5, CSS3, JavaScript (D3.js for Interactive Graph Visualizations)
-
Graph Engine: NetworkX for
$O(V + E)$ cycle detection efficiency.
To ensure eligibility and avoid disqualification, CypherRing meets all mandatory performance metrics:
| Processing Time | ≤ 30s for 10K transactions | ✅ Optimized | | Precision | ≥ 70% (Minimizing false positives) | ✅ Target Met | | Recall | ≥ 60% (Catching maximum fraud rings) | ✅ Target Met | | JSON Schema | Exact line-by-line field matching | ✅ Validated |