Wallet and market ingestion
BSC wallet activity, SOL route and pool context, ETH pool discovery, token metadata, and liquidity evidence.
Technology stack behind the Qinlong System.
Qinlong is built as separate research lanes, not one black box. Ingestion, candidate creation, route checks, validation logs, AI notes, and dashboard health are stored separately so failures can be traced.
Qinlong architecture
Risk-aware data flow
Signal normalization
Candidates are normalized with chain, source, wallet or pool context, buy pressure, FDV, liquidity, and route state.
Rule attribution
Each pass or rejection is tied to explicit chain-specific conditions such as source quality, liquidity, cooldown, or route evidence.
Validation storage
Research entries, exits, blocker reasons, and dashboard summaries are stored for later review.
Observability
Process health, scanner freshness, reporter status, validation state, and dashboard availability are monitored.
Operational boundaries
Research alerts, validation records, chain-specific diagnostics, and execution-readiness decisions remain separate by design.
Where AI is used.
AI reads the recorded evidence and writes short review notes. It is not allowed to turn a candidate into a live order or override chain rules.
Signal summarization
Raw events are condensed into a candidate note: source, chain, route state, missing evidence, and current status.
Blocker classification
The system labels common blockers such as weak source, missing route evidence, thin validation sample, or liquidity uncertainty.
Report generation
Reports turn dashboard state into a short summary for internal review or investor walkthroughs.
Strategy review
Historical validation records are compared for pattern review; old observations are not treated as current readiness proof.
Current automation boundary.
The current production boundary is research and review. Qinlong records candidates, quoteability diagnostics, validation records, and blocker reasons; it does not currently submit live orders automatically.
Future controlled automation is a roadmap consideration only after stronger evidence gates, operational controls, and risk boundaries are in place.
Security and data boundary.
Qinlong's public website is informational. It does not collect wallet credentials, signing material, exchange passwords, or API credentials. Product work is focused on research records, operational logs, and records that can be reviewed later.
No custody
Qinlong is not presented as a custody product and does not ask website visitors to transfer assets.
No website credential intake
The public site uses company email for contact and does not provide forms for sensitive access material.
Separated records
Evidence records, blocker reasons, operational health, and AI notes stay separated for review.
Why the system needs reliable compute.
Qinlong is not a static research note. It needs to collect changing market data, run repeated route checks, store validation records, and keep operational health visible while volatile markets move.
01
Frequent ingestion
Signals and liquidity context lose value quickly, so the system needs repeatable data collection rather than one-off snapshots.
02
Evidence storage
Each pass, rejection, route failure, and validation sample needs a durable record that can be reviewed later.
03
Operational visibility
Scanner freshness, route checks, validation state, and dashboard health must be observable before any scale-up.