
Leading B2B Marketplace: Decade-old B2B marketplace stabilized and re-architected. 70% API latency cut. Teams shipping features again after years of compounding technical debt.
Snapshot
Client
Leading B2B Marketplace
Industry
Retail & Ecommerce - B2B Surplus & Liquidation Marketplace
Geography
Solingen, Germany
Size
Boutique operator: Described as a market leader in the German surplus marketplace segment for over 10 years
Challenge
Legacy platform stabilization, architectural refactoring, and capability restoration
Services
Backend refactoring & containerization, seller UI modernization, search & caching optimization, multilingual support via DeepL, security hardening (AWS WAF), living documentation
Duration
Ongoing
Team
Not specified
70%
reduction in API latency across critical seller workflows
70%
reduction in API latency across critical seller wo
99.9%
uptime sustained post-recovery, supporting enterpr
25%
delivery velocity
Download this case study as a PDF
Shareable leave-behind · auto-generated · always up to date
Client Context
The client is the company behind a leading B2B online marketplace in Germany for surplus goods, excess inventory, and liquidation stock. Operating for more than a decade, the enterprise has run the platform to connect wholesale buyers and sellers across multiple product categories. The marketplace model demands high availability, reliable search, and smooth seller workflows, which is a combination that becomes increasingly difficult to maintain when the underlying technology accumulates years of unmanaged growth.
The Challenge
Years of rapid growth had transformed the marketplace into a high-performing commercial asset built on a foundation that had quietly become unmanageable. The original development team had rotated out, and codebase ownership was fragmented. Business logic had been embedded across the architecture without documentation, meaning that changes which appeared minor could trigger unpredictable failures in production. The system worked, but no one on the current team could confidently explain why, or predict what would break if any part of it changed. Three specific failure modes were compounding: No clear service ownership: Knowledge silos had become a structural risk. When something broke, the team had to reverse-engineer the system rather than consult documentation or a responsible owner. High change risk: Even routine UI updates or API modifications carried genuine production risk. The cost of every deployment was anxiety, and eventually, deployment frequency dropped to protect stability. Growing demand, shrinking system trust: As more sellers and buyers came onboard, the platform’s brittleness became more exposed. Performance degraded under load, and new feature requests stalled because the team lacked confidence in the system’s behavior. The client needed the platform stabilized, made structurally legible, and unlocked for development, all without interrupting a live marketplace that generated ongoing revenue.
The Approach
Gradion approached the engagement through a combination of reverse engineering, architectural refactoring, and targeted functional modernization, working in parallel rather than sequentially to minimize time to stability. Backend rebuild and containerization: Gradion migrated the legacy backend to a clean, modular architecture with explicit domain boundaries and observable logic flows. Containerization made individual services deployable and testable in isolation for the first time. Seller UI modernization with Laravel and Bootstrap: The seller-facing interface was rebuilt using Laravel and Bootstrap, replacing brittle legacy front-end code with responsive, maintainable interfaces. The result required no onboarding for existing sellers and eliminated a category of UI-related support requests. Search and caching architecture: Gradion restructured the search layer using Elasticsearch and implemented data caching across high-traffic queries, producing near-instant results under the platform’s peak load conditions. Multilingual product management via DeepL API: Automated translation was integrated directly into the seller listing workflow, enabling product descriptions to be published across language markets without manual intervention. Security hardening with AWS WAF: Request-level protections were added and infrastructure permissions were audited and cleaned up, producing a materially stronger security posture with clear ownership of cloud policies. Living documentation: Gradion generated comprehensive, up-to-date technical documentation covering all refactored services, permanently eliminating the knowledge-silo risk that had made the system unmanageable.
70%
Reduction in API latency across critical seller workflows
99.9%
Uptime sustained post-recovery, supporting enterprise-level service commitments
2x
Team expansion: The client expanded Gradion's scope and team size
The Results
The engagement converted a platform that the internal team feared touching into one they could actively develop against. The hard metrics tell the stabilization story: 70% reduction in API latency: This was achieved across critical seller workflows. 99.9% uptime sustained post-recovery: This supports enterprise-level service commitments to marketplace participants. +25% delivery velocity: With technical confidence restored, product teams began shipping new features that had been blocked. 2x team expansion: The client expanded Gradion’s scope and team size after the initial engagement, transitioning from firefighting mode to proactive strategic development. Knowledge-loss risk eliminated: The living documentation asset alone eliminated the structural risk that had created the crisis in the first place, making the platform resilient to future team changes.
Services & Technology
Services delivered
- Backend refactoring and containerization
- Seller UI modernization
- Search architecture and caching optimization
- Multilingual product management integration
- Security hardening and infrastructure audit
- Living technical documentation
Technology stack
- Laravel (backend modernization)
- Bootstrap (seller UI)
- Elasticsearch (search layer)
- DeepL API (automated multilingual translation)
- AWS WAF (security and infrastructure hardening)
- Docker (containerization)
Engagement model
Rescue engagement transitioning to strategic development partnership
Talk to us about restoring a platform that's too fragile to touch.
Describe the system. We will assess it, stabilize it, and give your team back confidence.