Varpulis Real-World Scenarios

Five industry scenarios demonstrating Varpulis's key differentiators: native Kleene closure, Hamlet multi-query optimization, sub-millisecond latency, and 3-16x memory advantage.

Scenario	Industry	Key Differentiator	Document
Credit Card Fraud Detection	Financial Services	Kleene closure captures 5x more fraud patterns than greedy-matching competitors	VPL
Predictive Equipment Failure	Manufacturing / IoT	Multi-sensor correlation with explainable alerts	VPL
Insider Trading Surveillance	Capital Markets	Hamlet trend_aggregate monitors 50+ symbols at 100x throughput	VPL
Cyber Kill Chain Detection	Cybersecurity / SOC	Multi-stage attack detection with complete forensic trail	VPL
Patient Safety Monitoring	Healthcare	Sub-millisecond drug interaction detection across prescribers	VPL

Running the Scenarios

Each scenario has a .vpl (rules) and .evt (test events) file in tests/scenarios/.

# Run all CxO scenario tests
cargo test -p varpulis-runtime --test cxo_scenario_tests

# Run a specific scenario
cargo test -p varpulis-runtime --test cxo_scenario_tests cxo_fraud

# Run with a live event file
varpulis simulate tests/scenarios/cxo_fraud_detection.vpl \
    --events tests/scenarios/cxo_fraud_detection.evt

Benchmark Highlights

Metric	Varpulis	Traditional CEP	Advantage
Single-query throughput	6.9 M events/s	2.4 M events/s	3x
Multi-query (50 patterns)	950 K events/s	9 K events/s	100x
Memory per instance	10-54 MB	85-190 MB	3-16x less
Latency	Sub-millisecond	15-minute batch cycles	Real-time