Architecture Overview
AsterXBT’s architecture is designed for scalability, security, and ease of deployment. It consists of the following core components:
Agent Core: The central AI engine powered by multi-LLM consensus. Each LLM processes inputs independently, and a voting mechanism determines the final action. Self-learning is managed through a feedback loop utilizing historical data stored in a decentralized database, which is enhanced with sentiment data for more accurate market predictions.
Data Ingestion Layer: Aggregates data from multiple sources:
On-chain data from Aster, a Perpetual DEX (order books, liquidity pools).
Off-chain feeds: APIs for technical indicators, news aggregators, sentiment tools, and specialized sentiment providers.
Custom user inputs: Configurable via a web-based dashboard or CLI, including sentiment thresholds and source priorities.
Execution Engine: Interfaces with Aster mart contracts to execute perpetual trades. Supports high-frequency trading with low-latency connections, incorporating sentiment signals for timely entries and exits.
Last updated