BattleMetrics event modeling
Player join, update, leave, report, kick, death, warning, anti-hack, and RCON events can become structured evidence streams for moderation and analytics.
Research & Labs
Practical research for safer multiplayer operations.
TIRust research focuses on moderation systems, event ingestion, risk scoring, evidence retention, workflow design, and AI-assisted review that supports staff instead of replacing them.
Evidence pipelines for fair-play operations.
Anti-cheat and moderation systems need disciplined ingestion, correlation, scoring, and review. Strong enforcement should be explainable, appeal-aware, and backed by a durable evidence trail.
IngestBattleMetrics triggers, Server Armour reports, Arkan alerts, RustAdmin events, Discord reports, RCON logs.
CorrelateSteam IDs, aliases, IP/ASN, session duration, first seen, org history, related players, prior flags.
ScoreComposite risk with decay, thresholds, explainable factors, and separate confidence levels.
ReviewStaff queue, evidence summaries, notes, action history, appeal context, and audit logs.
False-positive control
The system should never treat one weak factor as proof. Composite scoring, human review, confidence labels, and evidence capture are the guardrails.
AI-assisted review
LLMs can summarize evidence packets, cluster reports, explain risk factors, draft staff notes, and identify missing context. They should not issue irreversible actions by themselves.
Public data tooling
Labs may include GIS/API discovery, infrastructure indexes, monitoring experiments, and other tools that support broader operational research.