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

# Architecture

> The architecture overview of AstraDEX including AI Agent, AsterDEX API integration, and Telegram Mini App infrastructure.

## Overview

AstraDEX is built as a modular system that combines decentralized trading infrastructure with intelligent AI-driven analysis.  

It leverages the **AsterDEX API** for on-chain trading operations while providing an interactive user experience through the **Telegram Mini App**.  

The system is designed to be **scalable, secure, and adaptive**, ensuring seamless interaction between users, smart contracts, and AI agents.

The overall architecture consists of three main layers:

1. **Frontend Layer** – Telegram Mini App interface  
2. **Application Layer** – AI Chat Agent, Trading Logic, and Middleware  
3. **Infrastructure Layer** – AsterDEX API, Market Data, and Smart Contracts

## Frontend Layer (Telegram Mini App)

The **Telegram Mini App** acts as the main user interface for AstraDEX.\
It allows users to access trading features, interact with the AI chat agent, and view real-time market data inside Telegram.

### Key Features

* Seamless access using Telegram login without additional registration
* Real-time chat-based trading commands and market queries
* Responsive design for both mobile and desktop Telegram clients

### Communication Flow

The Mini App communicates securely with the **AstraDEX Backend Gateway** via HTTPS APIs.\
Each request — such as placing an order or fetching market data — is validated and routed to the correct service (AI Agent, Market Data, or Trading API).

## Application Layer

This layer represents the **intelligent core** of AstraDEX.\
It manages user interaction, AI-based decision-making, and integration with the trading engine.

### AI Chat Agent

The AI Chat Agent understands user input and responds with relevant trading insights.\
It can:

* Analyze current market trends
* Generate short-term forecasts
* Provide educational explanations about trading concepts
* Offer personalized trading suggestions

It uses **Natural Language Processing (NLP)** and **machine learning models** trained on trading data to interpret intent and generate meaningful outputs.

### Market Data Service

This service retrieves and caches live data from the **AsterDEX Market API**, including:

* Real-time prices and tickers
* Order book depth
* Historical Kline (candlestick) data
* Volume and volatility statistics

Cached data helps the AI respond faster and reduces latency for Telegram users.

### Trade Execution Service

This module interacts directly with the **AsterDEX Trading API** to execute user commands such as:

* Placing buy or sell orders
* Managing open positions
* Checking balances and leverage levels
* Tracking performance and historical trades

All requests are authenticated with secure API keys or wallet-based signatures.

### AI Strategy Engine

The AI Strategy Engine performs advanced market evaluation by combining:

* Technical analysis indicators (EMA, RSI, MACD, etc.)
* Machine learning–based sentiment analysis
* Short-term trend forecasting

It provides actionable insights like “Buy the dip” or “Wait for confirmation,” helping users make informed trading decisions.

## Infrastructure Layer

The Infrastructure Layer connects AstraDEX to the AsterDEX blockchain ecosystem.\
It focuses on security, scalability, and data reliability.

### AsterDEX API Integration

AsterDEX APIs are the backbone of AstraDEX’s trading functionality.\
They provide:

* Market Data endpoints for both spot and futures markets
* Trade Execution endpoints for limit, market, and conditional orders
* User Data endpoints for portfolio and balance management

All requests use **HMAC-SHA256 signatures** to verify authenticity and integrity.

### Smart Contracts

All on-chain operations are handled by **verified AsterDEX smart contracts**.\
These contracts control:

* Order matching and execution
* Liquidity pools and market-making
* Fee distribution and settlement
* Margin and leverage management

Because they are **non-custodial**, users retain full control of their assets throughout every transaction.

### Database and Caching

AstraDEX uses a hybrid data layer for performance:

* **PostgreSQL / MongoDB** for persistent data (user sessions, preferences, logs)
* **Redis** for real-time caching of market data and AI query results

This combination ensures low-latency performance and smooth scalability.

## Data Flow Summary

1. **User Interaction** – The user sends a message or command through the Telegram Mini App.
2. **Request Handling** – The backend identifies whether it requires AI analysis, trading execution, or market data.
3. **API Requests** – AstraDEX communicates with AsterDEX APIs for live data or trade operations.
4. **AI Processing** – The AI module analyzes the market and returns insights or trading signals.
5. **Response Delivery** – The Telegram Mini App displays the analysis, trade confirmation, or educational response.

## Scalability and Security

AstraDEX is built on a **microservices architecture**, allowing independent scaling of chat, trading, and AI components.\
This ensures high availability even under heavy user traffic.

### Security Highlights

* End-to-end encryption between Telegram and AstraDEX servers
* JWT-based authentication for user sessions
* HMAC-signed API communication with AsterDEX
* Rate limiting and WAF protection for abuse prevention
* Publicly verified smart contracts with audit transparency