Skip to main content

Kaldrop: Dropshipping Automation

A product-import platform that turns supplier listings into clean, ready-to-manage product records — with background import jobs and a dedicated scraping microservice built for reliability.

Visit website
  • Full Stack Development
  • System Architecture
  • Web Scraping
  • Automation
The Kaldrop dashboard showing product import activity.

Overview

Kaldrop is a dropshipping/product-import platform that helps sellers convert messy supplier listings (AliExpress, eBay, and similar marketplaces) into structured, normalized products inside a management dashboard.

The goal is simple: import faster, stay consistent, and scale catalog building — while keeping scraping complexity isolated from your core business backend.

What it solves

  • Sellers waste hours copying and cleaning product data.
  • Supplier pages are inconsistent, noisy, and change frequently.
  • Imports need to be repeatable, trackable, and resilient when scraping is temporarily unavailable.

Kaldrop turns the import flow into a reliable pipeline:

Find product → Import → Normalize → Manage & publish

Key features

Product import (multi-mode)

  • By URL / Supplier ID Submit supplier URLs or shorthand IDs, scrape the listing, normalize the data, then store it as a clean internal product record.
  • By CSV Ingest CSV rows, group variants correctly, and create products in bulk.
  • Import Jobs Imports run asynchronously with progress tracking and clear state transitions: pending → processing → completed / failed

Scraping service toggle + resilience

The platform is designed to keep the core backend stable even when scraping is not.

  • Scraping can be enabled/disabled via environment and database System Settings.
  • If scraping is down/disabled, URL import jobs can remain pending and auto-resume once scraping becomes available again.

Authentication & authorization

  • JWT authentication enforced via Express middleware (Authorization: Bearer ...).
  • Role-based access for Admin, User, and Affiliate flows.
  • Protected dashboard and admin routes.

Admin/member management (tRPC)

  • Admin-only member listing, creation, updates, and role changes.
  • tRPC endpoints + Zod validation for typed, safe mutations.

Architecture

Kaldrop is split into three cooperating services so each layer can scale independently:

  • Frontend (kaldrop-next) Next.js dashboard + admin UI, route protection, and API wrappers.
  • Backend (Express + TypeScript) Core SaaS features (accounts, products, imports, plans/subscriptions/affiliate) persisted in MongoDB via Mongoose.
  • Scraping microservice (mini-scraping-server) Puppeteer automation + Redis caching and TTL strategy, exposed through a simple API surface.

Reliability & performance highlights

  • Decoupled scraping keeps your main API responsive and reduces operational risk.
  • Redis caching reduces repeated scraping and stabilizes import throughput.
  • Job orchestration ensures imports are retry-friendly, observable, and resilient to downtime.

Tech stack

  • Frontend Next.js 14, React 18, Redux Toolkit, React Query, Axios, NextAuth, MUI, styled-components, TailwindCSS, tRPC client.
  • Backend Node.js, TypeScript, Express, MongoDB, Mongoose, tRPC server, Zod, Swagger (dev).
  • Scraping Service Express (ESM), Puppeteer + stealth plugin, ioredis caching, proxy support.