$ cat ./suryamasniagajaya/README.md
PT. Suryamas Niaga Jaya - Corporate Web Portal & AI Assistant
The PT. Suryamas Niaga Jaya project is a comprehensive corporate web application designed to digitize and streamline the B2B distribution of industrial chemicals and laboratory equipment. Built to serve both clients and internal administrators, the platform features a deeply hierarchical product catalog (Field -> Subcategory -> Product) to neatly organize thousands of complex technical items. Beyond standard catalog browsing, the platform integrates a custom AI Chat Assistant. By leveraging a vector database (Qdrant) and Large Language Models, the assistant provides semantic search capabilities, allowing customers to query for products using natural language, receive intelligent recommendations, and quickly connect with sales representatives. Additionally, the application includes a full-fledged CMS for managing blog posts (via TipTap/CKEditor), company partnerships, and dynamic content, wrapped in a premium, responsive user interface.
Case Study & Content
The Challenge
PT. Suryamas Niaga Jaya manages a massive inventory of industrial chemicals and laboratory equipment. Their previous product discovery process relied heavily on conventional methods, creating bottlenecks in customer support as clients struggled to locate specific formulations, equipment, or product variants.
The company needed a scalable digital platform capable of organizing its extensive product catalog while helping customers discover relevant products without immediately depending on human sales representatives.
The Solution
We engineered a modern, high-performance web application using Next.js, backed by a robust MySQL database managed through Prisma ORM. The platform combines an organized product management system with an AI-powered product assistant to improve product discovery and customer engagement.
Key Features Implemented
Hierarchical Catalog System
Designed a structured relational catalog that organizes products into dedicated Fields and Subcategories, enabling intuitive navigation for customers while simplifying product management through a custom administrative dashboard.
AI-Powered Product Assistant (RAG Implementation)
Implemented a Retrieval-Augmented Generation (RAG) pipeline that enables the AI assistant to understand customer questions and recommend relevant products based on semantic search instead of simple keyword matching.
- Vectorization – Product information, including descriptions, specifications, and pricing, is converted into vector embeddings.
- Semantic Retrieval – User queries are matched against a Qdrant vector database to retrieve the most contextually relevant products.
- LLM Response Generation – Retrieved product information is supplied to a Large Language Model to generate natural and accurate product recommendations in Indonesian.
- Graceful Fallback – When no suitable product is found, the assistant recommends contacting the sales team through an integrated WhatsApp channel.
Content Management & SEO
Integrated a complete content management system for publishing articles and industry insights using Tiptap Editor. The application leverages the Next.js App Router with server-side rendering (SSR) to improve search engine visibility and deliver fast page loading performance.
The Results
The completed platform represents a significant step forward in PT. Suryamas Niaga Jaya's digital transformation, providing a modern, scalable, and user-friendly B2B experience.
Powered by Tailwind CSS and Framer Motion, the interface delivers a polished and professional user experience. Meanwhile, the AI-powered product assistant reduces the initial workload of customer service representatives by guiding customers toward relevant products, improving engagement and creating a more efficient path from product discovery to customer inquiry.
$ ls ./screenshots/