> For the complete documentation index, see [llms.txt](https://docs.lyslabs.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lyslabs.ai/architecture.md).

# Architecture

The LYS platform is built from the ground up to deliver ultra-fast, structured, and AI-ready blockchain data. This architecture enables data ingestion, decoding, enrichment, aggregation, and delivery - all within milliseconds.

Here’s how the full system is structured:

```
LYS Data Stack
──────────────

DECODERS ─────┐
              │
              ├──> ETL: LYS Core
              │     └── Decodes and structures Solana blockchain events
              │
              ├──> Real-Time Aggregations
              │     └── Triggers: ATH, volume spikes, bundle detection
              │     └── Output: Live WSS streams for bots, dashboards
              │
              ├──> Historical Aggregations
              │     └── Indexed event storage for backtesting, PnL, token history
              │
              ├──> Solexys ETL
              │     └── Specialized decoding for agent insights and AI context
              │
              ├──> Solexys Blockchain API
              │     ├── Enriched token data for AI agents
              │     └── Solexys Frontend queries live metrics and wallet behavior
              │
              ├──> Solexys Off-Chain API
              │     └── Handles chat history, user-defined strats, watchlists
              │
              └──> Raw Data API & WebSocket
                    └── Raw decoded Solana block data in real time

```

### Data flow summary

1. **Raw ingestion**\
   Full Solana blocks ingested from Geyser + validator infrastructure.
2. **Protocol-specific decoding**\
   Custom parsers extract true on-chain intent from transactions.
3. **Event normalization**\
   Unified schema output: swaps, transfers, votes, launches, etc.
4. **Aggregation engines**\
   Real-time summarization (e.g volume, ATHs) + historical indexing.
5. **API layer**\
   REST + WebSocket endpoints for raw and enriched data delivery.
6. **Agent layer (Solexys)**\
   NLP-driven strategy interface backed by real-time intelligence.

{% hint style="info" %}
This architecture ensures that whether you're building trading bots, training AI agents, or doing advanced research - LYS delivers structured data with **speed, context, and precision.**
{% endhint %}