> ## Documentation Index
> Fetch the complete documentation index at: https://ducs.surchi.xyz/llms.txt
> Use this file to discover all available pages before exploring further.

# How SURCHI Works: From AI Signal to On-Chain Execution

> Learn how SURCHI ingests on-chain data, runs AI inference through the Sentinel Core, and executes strategies autonomously on Solana.

SURCHI operates as a continuous intelligence loop — ingesting raw blockchain data, processing it through specialized AI models, generating actionable signals, and optionally executing strategies on-chain without requiring manual intervention. The entire pipeline runs autonomously, with you remaining in control of the risk boundaries the Sentinels operate within. Here is how each stage of that loop works.

## The Intelligence Pipeline

<Steps>
  <Step title="Data Ingestion">
    SURCHI's **Data Pipeline** continuously ingests a broad spectrum of real-time signals from across the Solana ecosystem and beyond:

    * **On-chain transactions** — every swap, transfer, and contract interaction relevant to monitored assets
    * **DEX order books** — real-time order flow and depth data across Solana's major decentralized exchanges
    * **Liquidity pool depths** — moment-to-moment pool composition changes across AMMs and concentrated liquidity protocols
    * **Whale wallet movements** — activity tracking for known large holders and high-impact wallets
    * **Social sentiment signals** — aggregated analysis of market sentiment from public discourse channels

    All data is normalized and timestamped before being passed downstream for analysis.
  </Step>

  <Step title="AI Analysis">
    The **Sentinel Core** processes the raw ingested data through a suite of specialized AI models, each tuned for a distinct analytical task:

    * **Alpha Sentinel** evaluates market opportunity — identifying momentum, divergence, and early accumulation patterns that precede significant price movements
    * **Liquidity Sentinel** monitors pool health and risk — detecting early signs of liquidity deterioration, abnormal withdrawal patterns, and conditions that historically precede adverse events
    * **Execution Sentinel** formulates action plans — determining optimal entry and exit parameters, sizing, timing, and routing based on current market microstructure

    The three Sentinels share a common context layer, allowing cross-model intelligence to inform each agent's outputs.
  </Step>

  <Step title="Strategy Generation">
    Processed intelligence is surfaced as **actionable signals, alerts, and execution recommendations** delivered through the SURCHI interface. You receive:

    * High-confidence opportunity alerts from Alpha Sentinel
    * Liquidity risk warnings and pool health summaries from Liquidity Sentinel
    * Proposed execution strategies with projected parameters from Execution Sentinel

    You can also bypass waiting for automated signals and interact directly with the Sentinels using the **Natural Language Interface** — describing your intent in plain English to receive an immediate analysis or trigger an execution plan.
  </Step>

  <Step title="Autonomous Execution">
    When you have configured the Execution Sentinel with authorization, SURCHI executes **on-chain transactions on Solana** with sub-second speed and precision. The Execution Sentinel:

    * Selects optimal DEX routing and transaction timing
    * Applies slippage protection and gas optimization
    * Monitors transaction status and handles retries if necessary
    * Operates strictly within the risk parameters, position limits, and execution permissions **you have defined**

    No execution occurs outside your authorized boundaries.
  </Step>

  <Step title="Continuous Learning">
    Every execution — and every outcome — feeds back into SURCHI's AI models as training signal. The protocol continuously refines its:

    * Market pattern recognition
    * Liquidity risk models
    * Execution quality and routing optimization

    Over time, the Sentinel Core becomes increasingly accurate and adaptive to evolving market conditions.
  </Step>
</Steps>

## The Three Sentinels

Each Sentinel is a purpose-built AI agent with a specialized role within the SURCHI ecosystem. Together, they cover the full intelligence-to-execution lifecycle:

| Sentinel                                            | Role                 | Focus                                                        |
| --------------------------------------------------- | -------------------- | ------------------------------------------------------------ |
| [Alpha Sentinel](/sentinels/alpha-sentinel)         | Market Intelligence  | Opportunity detection, trend analysis, sentiment correlation |
| [Liquidity Sentinel](/sentinels/liquidity-sentinel) | Liquidity Protection | Pool health monitoring, risk detection, position safety      |
| [Execution Sentinel](/sentinels/execution-sentinel) | Autonomous Execution | On-chain strategy execution, routing, risk enforcement       |

For a deeper look at each agent, see the [Sentinels Overview](/sentinels/overview).

<Tip>
  You stay in control at all times. Set your own risk limits, position sizes, and execution permissions before the Execution Sentinel can act. The Sentinels operate as your agents — never beyond the boundaries you establish.
</Tip>