← Back to system

System Architecture

A high-concurrency microservices ecosystem built for professional data management. Focusing on reliability, distributed state management, and real-time feedback loops.

01. Core & Orchestration

App Core Service

PHP 8.4 / Yii 3 / RoadRunner

The backbone of ResumeGo. Manages business logic, auth, and complex data schemas. Uses RoadRunner to stay memory-resident, eliminating traditional PHP overhead.

Workflow Orchestrator

Temporal.io

Governs the distributed 'Resume Generation' lifecycle. Guarantees eventual consistency and execution reliability through durable state tracking.

02. Specialized Workers

PDF Generator

Go (Golang)

Compute-intensive worker. Ingests structured JSON to produce pixel-perfect PDF documents using highly concurrent rendering routines.

AI Enhancement Service

Go / OpenAI API

Handles semantic text optimization. Implements streaming and asynchronous field completion for professional experience and summaries.

03. Messaging & Real-time

Event Broker

RabbitMQ

Internal communication bus. Features persistent exchanges and dedicated retry queues to ensure notification delivery to the WebSocket layer even after user disconnects.

Notification Hub

Node.js / WebSockets

Ultra-lean socket bridge. Consumes RabbitMQ signals and pushes instant feedback to the React frontend.

04. Data & Storage

Primary Database

PostgreSQL + Redis

Relational integrity for core entities (Postgres) combined with an ultra-fast Redis layer for session state and dynamic caching.

Object Storage & I18n

RustFS (S3) + Tarantool

RustFS provides high-performance file persistence. Tarantool (Lua) serves localization data directly from RAM with sub-millisecond response times.

05. Frontend Environment

Application Framework

Next.js 16 / React 19

Utilizing the latest App Router paradigms. Leveraging React Server Components (RSC) for performance and streaming for high-impact descriptions.

State & UI System

Zustand / shadcn/ui / Tailwind CSS

Lightweight state management via Zustand for complex forms. UI built with shadcn/ui primitives and Tailwind CSS for zero-runtime overhead styling.

Developer Note

"Primary expertise lies in Backend Engineering. The frontend implementation was facilitated by AI generation tools to streamline the development of the UI layer while I focused on the core system architecture."

Observability Stack

Logs & Metrics Collection

Telegraf agents collect system metrics, feeding InfluxDB for time-series analysis and OpenObserve for high-performance log indexing.

Visualization & Edge

Centralized monitoring dashboards via Grafana. The entire cluster is unified under Caddy Proxy for Layer 7 routing and SSL termination.

Architecture version: 2026.4.PoC

Status: Active Development