How It Works
Speed without structure is noise. Here's the actual mechanics behind the system.
Full Launch Surveillance
The engine monitors every token deployed on Pump.fun. No manual filtering. No curated lists. No selective indexing. The moment a token is created, our system registers it.
- → Timestamped at first detection
- → Indexed by mint address
- → Linked to deployer wallet
- → Logged into our internal scoring engine
The timestamp in the app is our detection time — not blockchain creation time. Detection latency defines opportunity windows.
Proprietary Multi-Layer Assessment
Every token flows through our internal scoring model. The formula is proprietary. The structure is layered. Each token accumulates a composite score that determines its classification tier. The model recalculates continuously as new data flows in. Tokens are not judged once — they are judged repeatedly.
The Three-Tier Classification System
Earliest classification beyond raw detection. Token has passed baseline integrity checks and shows early structural signals worth monitoring. Warm does not imply quality — it indicates structural viability that meets minimum thresholds.
Tokens graduate to Hot when the probability stack strengthens — cleaner distribution, sustained volume aligned with unique wallet growth, reduced concentration risk. Hot tokens show characteristics that historically precede larger moves, but remain fluid.
The highest classification tier. These tokens meet deeper structural standards and unlock expanded data visibility inside the app — full security analysis, holder concentration, social links, and external verification links. Not every token reaches this tier. Classification is not permanent. It is dynamic.
Dynamic Removal Protocol
A token that qualifies today may fail tomorrow. Our system continues evaluating after classification. If structural integrity weakens, the classification adjusts. If the probability stack collapses, the token is removed entirely.
Removal can occur due to liquidity instability, sudden concentration shifts, deployer anomalies, or volume pattern distortion. We do not keep tokens displayed for optics. We remove what no longer qualifies.
Real-Time Intelligence Flow
The backend operates continuously. Scoring recalculations occur as new on-chain data becomes available. Our system does not rely on manual refresh cycles — it tracks state transitions.
- ✓ Tier changes happen automatically
- ✓ Data panels update dynamically
- ✓ Removals occur without delay
📬 Telegram Alert Layer
Speed matters. We push high-signal updates through our Telegram channel when tokens enter significant structural states.
Join Telegram Channel