GMS Dashboard Pro is an enterprise-grade E-commerce Intelligence Platform tailored for high-volume Amazon India operations. This document provides an exhaustive technical deep-dive into the platform's architecture, data models, automated pipelines, and intelligence layers.
The platform is designed around a Service-Oriented MERN Stack, prioritizing data integrity, high-speed aggregations, and automated marketplace synchronization.
graph TD
subgraph "Frontend Layer (React 19)"
UI[Pro Dashboard UI] --> Store[Context API / State]
Store --> Components[ApexCharts / Tabler Icons]
end
subgraph "API Gateway (Express.js)"
API[API Router] --> Middleware[Auth/RBAC Middleware]
Middleware --> Controllers[Module Controllers]
end
subgraph "Automation & Intelligence"
Controllers --> Sync[Octoparse Sync Service]
Sync <--> Jobs[Self-Healing Sync Loops]
Controllers --> AI[Perplexity / NVIDIA NIM SDK]
end
subgraph "Persistence Layer"
Controllers --> MongoDB[(MongoDB Atlas)]
MongoDB --> Models[Mongoose Schema Layer]
end
gms-dashboard/
βββ backend/ # Node.js / Express Server
β βββ config/ # DB & App Configuration
β βββ controllers/ # Business Logic (30+ Controllers)
β βββ cron/ # Scheduled Task Orchestration
β βββ middleware/ # Auth, Upload, RBAC Logic
β βββ models/ # Mongoose Schemas (33 Models)
β βββ routes/ # REST API Definition
β βββ services/ # External API Wrappers (Octoparse, NIM)
β βββ uploads/ # Local Asset Storage (ASIN Images)
β βββ utils/ # Calculation & Validation Helpers
βββ src/ # React 19 Frontend
β βββ components/ # Premium UI Component Library
β βββ contexts/ # Global State Management
β βββ hooks/ # Reusable Logic (API, UI)
β βββ pages/ # Module-Specific Views
β βββ styles/ # Zinc Design System Utilities
β βββ utils/ # Frontend formatting & UI logic
βββ public/ # Static Assets
βββ package.json # Core Dependencies (ApexCharts, Lucide, etc.)
The system utilizes MongoDB Atlas for document-oriented storage, with heavy optimization for time-series data (Ads) and relational-like tracking (ASINs).
The Asin model is a sophisticated document that tracks 110+ data points including:
- Identity: ASIN, SKU, Seller ID, Brand, Category.
- Market State:
currentPrice,bsr,rating,reviewCount,soldBy. - Historical Arrays:
history: Daily snapshot of Price/BSR/Rating.weekHistory: Week-over-week performance metrics for trend analysis.
- Intelligent Wrappers:
lqsDetails(Listing Quality Score) andfeePreview(FBA Profitability).
Optimized for daily/monthly performance tracking:
- Core KPIs: Spend, Sales, Impressions, Clicks, Orders.
- Attribution Logic: Automated calculation of ROAS, ACoS, CTR, and AOV.
- Indexing: Partial unique indexes ensure zero data duplication at the
ASIN+Date+reportTypelevel.
Supports the AI-driven strategy layer:
Goal: High-level business intent (e.g., "Improve Margin").Action: Decomposition of goals into trackable tasks withTime to CompleteandStatusloops.
The redesigned ASIN interface provides a "Mission Control" experience:
- Horizontal Product Intelligence: A top-bar summary housing all vital stats + live Buy Box monitoring.
- ApexCharts Triple-Stack:
- Price History: Indigo area chart with gradient fills.
- BSR Trend: Emerald line chart with Reversed Y-Axis (Top ranking visibility).
- Rating Progression: Amber line chart for long-term customer sentiment tracking.
- Smart Filtering: Integrated range selectors (7D, 30D, 90D, All) that slice through historical arrays in real-time.
The octoparseAutomationService manages complex web-scraping lifecycles:
- Self-Healing Sync: When local data deviates from marketplace reality, the system automatically queues a background scrape.
- Concurrent Orchestration: Manages 100+ concurrent seller tasks with automated retry and fallback logic.
- Data Reconciliation: Ingests disparate marketplace results into the unified
Asinhistory layer.
- SD3 Workflow: Automatically generates high-quality lifestyle images for ASINs failing the LQS image count threshold (< 7).
- Automation: Triggered directly from the ASIN Manager when "Image Optimization" is required.
- Decomposition: Uses advanced LLM prompts to split vague intentions into 4-week execution plans.
- Validation: Ensures every task generated is measurable against marketplace KPIs (ACoS, ROAS).
GET /api/asins: Unified list with seller filtering and aggregation.PATCH /api/asins/:id: Partial updates for LQS and marketplace status.POST /api/asins/sync: Manual trigger for the Octoparse orchestration.
POST /api/upload/ads-data: Complex CSV parser with Case-Insensitive Header Mapping supporting standard Amazon Advertising reports.
- Production-grade
.envincludes secrets for Clerk Auth, Keepa SDK, Octoparse API, and NVIDIA NIM.
- MongoDB Indexing: Compound indexing on
asinCode + selleranddate + reportTypeensures sub-100ms query performance on 1M+ records. - Vercel Edge Rendering: The frontend is optimized for zero-latency loading of massive DataTables.
Built on a bespoke design system focusing on Clarity and Precision.
- Typography: Inter (UI) / Outfit (Header) for maximum readability.
- Grid: 24px unified corner-radius system.
- Effects: 12px Backdrop Blur (Glassmorphism) for focused detail modals.
Β© 2026 Easysell Projects. Distributed under the ISC License. Confidential and Proprietary. All Rights Reserved.