Project overview
Schip & Ster is a Netherlands-based premium lighting brand selling designer lamps, fixtures, and accessories through its online store. They wanted to move from a traditional webshop to an intelligent ecommerce platform that could automate operations, improve customer experience, and unlock growth in the Dutch market.
I led the end‑to‑end design and development of their new AI-powered ecommerce site, built with React, TypeScript, Node, and Redis, and integrated with Gemini, Twilio, and Sendcloud for logistics and communication automation. Within a few months of launch, the new platform helped Schip & Ster more than double their online sales while simplifying daily operations for the team.twilio+1
Business goals and challenges
Before this project, Schip & Ster faced several key challenges typical for fast‑growing ecommerce brands in the Netherlands:
-
Manual, slow product onboarding for hundreds of lighting SKUs.
-
Limited SEO optimization and content depth per product.
-
Fragmented shipping operations across multiple carriers.
-
Limited analytics to understand product performance and marketing ROI.
-
No integrated AI or automation layer to reduce repetitive work.
The business goals for the rebuild were:
-
Build a modern, scalable ecommerce platform tailored to the Dutch market and categories like pendant lights, floor lamps, and smart lighting.
-
Use AI to automate product creation, SEO, and content workflows.
-
Centralize shipping through a multi‑carrier platform (Sendcloud) with label generation and tracking.
-
Implement analytics that show what is selling, where revenue comes from, and how promotions like seasonal deals are performing.
-
Improve customer experience with better search, navigation, and communication.
Architecture and tech stack
The platform is designed as a modular, AI‑ready ecommerce system:
-
Frontend: React + TypeScript single‑page application for the storefront and admin dashboards.
-
Backend: Node.js (TypeScript) API layer serving product data, orders, AI endpoints, and integrations.
-
Caching and speed: Redis for caching product lists, category pages, and AI results to keep the store fast even under high traffic.
-
Shipping: Sendcloud API integration to connect to multiple European carriers, generate labels, and manage tracking from a single API.
-
AI layer:
-
Gemini for product enrichment, SEO text, and smart content generation.
-
Internal AI services for automatic product structuring and category mapping.
-
-
Communication: Twilio for notifications and transactional messaging (order confirmations, shipping updates, etc.).twilio
This stack gives Schip & Ster a flexible, future‑proof foundation that can evolve with new AI features and channels.
Key AI-powered features
1. AI-assisted product onboarding
For a catalog with thousands of lighting products, manual data entry is extremely slow. To solve this, I built an AI quick product add workflow:
-
Admin uploads product images and provides a very short description or a few bullet points.
-
The backend sends this context to Gemini to:
-
Generate a full product title and long‑form description optimized for Dutch lighting customers.
-
Suggest technical attributes like material, color, style, suitable room, and compatible light sources based on the image and text.
-
Propose bullet‑point selling features and usage ideas.
-
-
The system structures this data into the internal product schema and pre‑fills pricing, variants, and tags where possible.
This turns a process that used to take 10–15 minutes per product into a 1–2 minute review and publish flow, which is critical when managing categories like “hanglampen,” “vloerlampen,” and “wandlampen.”
2. AI-driven SEO and on-site content
To improve organic reach and category ranking, I implemented an AI SEO assistant:
-
For each product and category, Gemini generates meta titles, meta descriptions, and SEO‑friendly content blocks aligned with relevant search intent (e.g., “moderne LED hanglamp,” “vloerlamp goud”).
-
The system also suggests internal links to blog posts (for example, about price drops and design inspiration) and related products based on semantics, not just hardcoded rules.
-
Admins can accept, edit, or regenerate suggestions, keeping humans in control while drastically reducing content work.
Over time, this has helped Schip & Ster build richer category pages and blog posts around seasonal deals and design trends, supporting higher visibility and conversion.
3. Analytics and business insights
I built a custom analytics layer tailored to Schip & Ster’s operations:
-
Revenue breakdown by category (e.g., pendant lights, table lamps, floor lamps).
-
Performance of “Deals” and seasonal campaigns like spring offers and summer discounts.
-
Product‑level metrics: views, add‑to‑cart rate, conversion rate, return rate.
-
Shipping and fulfillment metrics derived from Sendcloud data:
-
Average delivery time.
-
Carrier performance and failed delivery rates.
-
These insights help the team decide which products to promote, when to run campaigns such as large percentage discounts, and which carriers or methods to favor in the Dutch market.
4. AI-enhanced customer experience
To improve the customer journey, the platform includes:
-
Smart search tuned for product categories, tags, and synonyms in Dutch (for example, mapping user queries to the right lamp types).
-
Product recommendations on bestsellers, seasonal offers, and high‑margin items.
-
AI‑generated blog content around lighting trends, price drops, and styling tips, tied directly to in‑stock products.
Twilio integration enables future expansion into proactive notifications such as low‑stock alerts, restock messages, and personalized campaigns via SMS or WhatsApp, all driven by the same analytics layer.
Shipping automation with Sendcloud
For a Dutch ecommerce brand shipping thousands of parcels, reliable, automated logistics is critical. I integrated the platform tightly with Sendcloud’s multi‑carrier API:
-
Orders from the webshop are automatically synced to Sendcloud.
-
The system creates shipping labels in one API call, no matter which carrier the customer chooses at checkout.
-
Tracking links and status updates flow back into the platform so customers can access them via their account, email, or messages.
-
Shipping rules and service selection are configurable per country, order value, and weight.
This eliminates manual label creation, reduces errors, and keeps delivery promises like “order before 22:00, delivered tomorrow” realistic and trackable for customers across the Netherlands.
Results and impact
Since migrating to the new AI‑powered platform:
-
Online sales for Schip & Ster have more than doubled, driven by better conversion, richer content, and smoother operations.
-
Product onboarding time per SKU has dropped dramatically thanks to AI‑assisted creation.
-
SEO and content automation have allowed the team to maintain extensive category and blog content without a large editorial team.
-
Shipping and fulfillment became more predictable and scalable through Sendcloud, improving customer satisfaction and reducing manual work.
For a Netherlands‑focused lighting brand competing in a mature ecommerce market, this combination of modern stack, AI automation, and logistics integration has become a real strategic advantage.
Tech summary table
Aspect Implementation
Frontend React + TypeScript SPA for storefront and admin dashboards. Backend Node.js (TypeScript) with REST APIs for products, orders, AI, and logistics. Data & caching Redis for caching product and category data, improving performance. AI services Gemini for product descriptions, attributes, and SEO content. SEO automation AI-generated meta tags, descriptions, and internal linking for categories. Shipping Sendcloud multi‑carrier API for labels, tracking, and carrier routing. Communication Twilio for transactional notifications and messaging. Analytics Custom dashboards for revenue, product performance, and campaign impact. Market focus Dutch ecommerce market and lighting vertical (Schip & Ster brand).