Engineering a High-Throughput B2B Wholesale & Retail Engine
How Skillycat solved complex customer pricing matrixes, offline tiering, and real-time inventory synchronization for Trademaxx.
Overview
Trademaxx operates across both high-volume B2B wholesale distribution and direct-to-consumer (B2C) retail. The core engineering objective was to establish a single digital ecosystem capable of dynamic pricing calculation without degrading catalog browsing performance.
Business Context
Prior to this system, wholesale buyers submitted manual purchase orders via email and telephone spreadsheets, leading to inventory discrepancies, delayed quotes, and high administrative overhead during peak supply chain windows.
The Core Problem
Key business and technical hurdles that needed a structured solution:
- B2B bulk pricing matrixes required real-time customer level qualification.
- Inventory synchronization had to prevent stock allocation overlaps between retail buyers and wholesale bulk orders.
- Checkout flows needed support for traditional invoice-based credit terms alongside instant credit card payments.
Strategy & Approach
Rather than forcing traditional monolithic ecommerce software to stretch across wholesale logic, we designed a headless architecture powered by Next.js and a modular API backend. This enabled instant client-side rendering for retail shoppers while utilizing cached server-side credit validation for logged-in B2B accounts.
UX & UI Perspective
Wholesale buyers prioritize order velocity over exploratory browsing. We introduced a quick-order grid layout where buyers can input multi-SKU variants directly into a matrix table, calculate volume breaks dynamically, and submit purchase orders in under 60 seconds.
Technical Architecture
We utilized Next.js App Router for optimized static generation of product detail pages, paired with edge-cached state management for authenticated user session pricing. PostgreSQL handles transactional integrity with row-level locks on stock reservation.
Key Implementation Challenges
CHALLENGE: Calculating custom tier pricing dynamically for over 10,000 SKUs without triggering database latency spikes. SOLUTION: Implemented a memory-cached tier rules engine that resolves customer discounts at session initialization, passing a signed token to frontend requests for fast inline recalculation.
Operational Outcome
Lessons Learned
B2B shoppers demand speed and utility over ornamental animations—quick matrix inputs drive higher conversion than standard product grids.
Caching authorization tokens with pricing rules significantly reduces database load during flash wholesale events.
Key Outcomes
Eliminated the manual PO email and spreadsheet workflow — all wholesale pricing, quoting, and checkout is now handled digitally through a single platform.
Dynamic B2B tier pricing resolved at session initialization, enabling full catalog browsing across 10,000+ SKUs with zero per-request database overhead.
Row-level PostgreSQL stock locks eliminated overselling between concurrent wholesale and retail order channels, achieving 99.9% inventory accuracy.
Future Improvements
- Automated ERP synchronization connector for instant PO matching.
- Multi-currency dynamic localized checkout for overseas regional distributors.
Related Services & Engineering Solutions
Related Projects Portfolio
Mukthika Jewellery
A premium online jewellery storefront designed for Mukthika Jewellery to showcase handcrafted gold, diamond, and bridal jewellery collections with secure online purchasing.
OneHand Embroidery
A luxury boutique e-commerce platform specializing in handcrafted, artisanal embroidery and custom bespoke apparel.
Building something similar in Kerala?
Have a Similar Challenge?
Discuss your technical requirements, architectural questions, or design needs with our engineering team.