Course content.
6 weeks (self-paced) · 48 hours · 6 lessons
On-Chain Data for ML
Sourcing and indexing transactions, events, and traces at scale · Feature engineering: contract age, holder distribution, liquidity patterns · Building labelled datasets from known exploits and verified contracts
Rug-Pull & Scam Detection
Supervised classifiers for honeypot and rug-pull contracts · Token-launch signals: liquidity lock, ownership renounce, hidden mints · Temporal features and deployer-reputation scoring
Wash Trading & Market Manipulation
Graph-based detection of circular trading patterns · Volume anomaly models and statistical tests · NFT wash-trading heuristics and marketplace-specific signals
Phishing & Address Poisoning
Classifying approval-phishing contracts by bytecode similarity · Address-poisoning detection via dust-transaction patterns · NLP models for scam-URL and social-media lure detection
Real-Time Deployment
Streaming pipeline: mempool listener → model → alert · Webhook and Telegram/Slack integrations for ops teams · Model drift monitoring and retraining triggers
Integration & Capstone
Plugging model output into Chainalysis, Arkham, or internal dashboards · Precision/recall tuning for security: cost of false negatives vs false positives · Capstone: deploy a working detector on a testnet data feed