Skip to main content

​
Introduction

At Neurobro, we aim to build the new standard for AI information in crypto. This goal comes hand in hand with the excellence of the platform and implementation of real tech.

​
Platform Microservices Architecture

The Neurodex Platform comprises four primary services:

UI (Client)

Next.js frontend for user interactions

Backend (API)

FastAPI service for request handling

Neurobro Agents

Core intelligence service

Background Services

Background workers to keep the platform running smoothly
Together, they form the platform, which allows real-time tracking of the market, data processing, analysis, and keeping it up to date, agent coordination, application layer with backend/client, databases, and more.

​
Underlying Services

Here’s the list of the underlying services, which make the platform work:

Message Broker

RabbitMQ

Relational Database

PostgreSQL + PgBouncer

Document Database

MongoDB

Cache

Redis

Vector Store

Qdrant

Graph Database

Neo4j

SSL security

HTTPS via Let’s Encrypt and Traefik

CDN

Cloudflare

Error Tracking

Sentry

Analytics

PostHog + Google Analytics

Monitoring

Prometheus + Loki + Grafana + Opentelemetry + Alertmanager

Deployment

Docker + Kubernetes

​
Web Application Architecture

Framework

Next.js with SSR optimization

Authentication

Privy Auth integration

Swaps

Rango Exchange API integration
The main focus is on the simplicity and usability of the client. We pay attention to the UI/UX and always iterate to make it better.
  • FastAPI
  • Database layer
  • Message routing
  • Major services integration
The backend is responsible for the core functionality of the platform. It handles the requests from the client, processes the data, and simply connects all the services together.
  • Containerized replicas
  • Workflow-based execution
  • Real-time streaming
  • Nevrons orchestration
There are 150+ Nevrons (swarm AI Agents) in total, each with a unique purpose and set of tools. Here are some examples:

Market Analysis

nevron23-26 - Market Sentiment Analysis

Technical Analysis

nevron31-33 - Technical Analysis

Research

nevron12-17 - Project Research (Focus on web, official data, publications, etc.)

On-Chain Scam Detection

nevron77 - On-Chain scam detection (Focus on suspicious onchain activity)
  • Background workers to keep the platform running smoothly
  • Data processing and analysis
  • Market data updates
  • User data updates
  • Notification system
  • Agent communication and Memory management
  • etc.

​
Scalability & Performance

With scale comes complexity. There are many moving parts, and we make sure that the platform is scalable and performant. First of all, we scale the Nevrons swarm and the so-called “Agents Service” based on the load of background processes and real-time user requests.

Load Distribution

RabbitMQ workload balancing

Auto-scaling

Kubernetes auto-scaling

Low Latency

RPC streaming of processing results

​
Additional Resources

Nevron Documentation

Learn more about the Nevron framework and how it empowers a swarm of 150+ AI Agents

Changelog

See what’s new in the Neurodex
For technical support or architecture discussions, join our developer community.