> For the complete documentation index, see [llms.txt](https://docs.dexlens.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.dexlens.io/how-it-works/signal-generation.md).

# Signal Generation

* **Data Collection**: Dexlens employs a network of APIs and intelligent crawlers to continuously ingest and update live market and social sentiment data from multiple sources.
* **Signal Scoring**: A multi-layered AI system processes the information using distinct modules:
  * **Volatility**: Detects abrupt price movements and instability
  * **Liquidity**: Observes bundle behaviors and trading activity
  * **Wallet**: Tracks wallet clustering and notable movement
  * **Narrative**: Analyzes social momentum, keywords, and narrative trends
* **Bundle Detection**: Identifies suspicious bundle concentrations or distributions
* **Distribution Scoring**: Evaluates wallet concentration to flag potential risks
* **Signal Emission**: When indicators meet key thresholds, Dexlens generates a sentiment alert published to X and displayed within the Terminal