Swarovski partnered with CLOUDSUFI and Google Cloud to consolidate 30+ data sources into a single cloud-native lakehouse, cutting migration time, eliminating data variance, and freeing engineers to innovate.
- Migrated 1,000+ data objects from dozens of sources into a single cloud-native Data Lakehouse on BigQuery.
- AI logic translation agents converted legacy ABAP and SAP BW code automatically, cutting weeks of manual work.
- Automated validation gates ensured every migrated record matched its legacy counterpart exactly.
From legacy systems to data-driven luxury
Swarovski has been in the brilliance business since 1895. Over 130 years, the Austrian brand has grown into a global name with more than 2,300 boutiques in 140+ markets. Its crystals show up in couture, jewelry, and home decor. The craft is old. The data problem was new.
By 2022, Swarovski’s business was running on dozens of disconnected systems. Sales data lived in one place. CRM data in another. Supply chain, ecommerce, and marketing each had their own version of the truth. Insights were slow. Reports didn’t agree. And the legacy code that ran pricing, inventory, and segmentation was locked inside ABAP and SAP BW — decades of rules that no one wanted to touch.
“We came from a time when everyone had their version of the truth. Our data structures were mainly designed for financial reporting in the last two decades, not for real-time personalization in the 2020s.”
Customer expectations were moving in the opposite direction. Luxury shoppers now expect a single experience across online, mobile, and in-store. They expect the brand to remember them, recommend the right piece, and follow through on service. That kind of moment needs live data — not a report that shipped yesterday morning.
Swarovski needed more than a tech refresh. It needed a complete rethink of how data was stored, trusted, and used across the company. The goal was simple to name and hard to build: a single foundation the whole company could trust — and that AI could build on.
One golden source of truth, built for AI
Swarovski partnered with Google Cloud and brought in CLOUDSUFI, a Google Cloud partner, to help run the migration at scale. The team moved more than 1,000 data objects across every major business system into a single cloud-native Data Lakehouse on BigQuery. CRM, ERP, ecommerce, and creative assets all landed in one place, for the first time.
The work turned on three ideas. AI logic translation agents read the old ABAP and SAP BW code and rewrote it into modern SQL on their own — so engineers didn’t have to spend weeks hand-converting legacy business rules. Automated validation gates ran every record through a legacy-versus-cloud check, and nothing shipped unless the two sides matched exactly. AI-driven semantic mapping lined up schemas across systems, so a “customer ID” in one source and a “client reference” in another could finally agree on what they meant.
CLOUDSUFI designed and ran this AI-Led Migration Factory as the implementation partner on the program, working alongside Swarovski’s data team and Google Cloud’s technology stack.
“Gen AI removed the entry barrier to AI. Today, everyone can be a creator, an analyst, or a designer — without code.”
The new foundation opened the door to Vertex AI and Gemini on top of BigQuery. Swarovski launched Génie, an internal gen AI portal that now serves more than 1,000 employees for contract review, translation into 20+ languages, creative asset generation, and early-stage product cost estimation. A Customer Data Platform on Looker unified touchpoints across ecommerce, CRM, marketing, and retail, so every team could work from the same single customer view.
Measurable impact across the business
The headline numbers tell a short story. The AI translation layer cut migration time by more than a third compared to manual re-coding. Automated validation gates delivered 100% accuracy between legacy and cloud-native outputs, with zero variance on financial, operational, and customer records. Automated documentation and lineage freed up about 40% of ops capacity — time that used to go to pipeline maintenance and now goes to new data products.
The deeper change is what the business can now ask. Financial reporting used to look backward. Today it blends live marketing, campaign, and even weather data into forward-looking forecasts. AI-personalized email campaigns see higher open and click-through rates. Campaign localization runs far faster thanks to AI-assisted translation and asset adaptation. Customer service workflows triage tickets and assist agents in real time, instead of waiting on a nightly batch.
“Luxury today is about relevance, timing, and emotional connection. With Google Cloud, we’ve built intelligent solutions that listen, learn, and adapt in real time. We communicate authentically and intelligently with our customers, while keeping the human connection and creativity at the core. That’s what it means to bring joy.”
Swarovski paired the platform with a responsible AI framework. Every new AI use case goes through an internal ethics and risk review that weighs data privacy, brand integrity, and security. In 2024, 77% of AI initiatives were evaluated and approved through that framework — enough signal for the company to keep scaling, with guardrails that actually hold.
What’s next: A future built on data, creativity, and AI
Swarovski is piloting the next wave of gen AI across the business — digital merchandising, virtual try-on, predictive product recommendations, and conversational agents that can help customers and employees alike. Each pilot plugs into the same lakehouse, the same customer view, and the same governance framework.
The goal isn’t to replace craft with code. It’s to let human creativity and AI-powered scale work side by side — so a 130-year-old luxury brand can keep bringing joy to customers it has never met, around the world.
CLOUDSUFI is a Google Cloud Premier Partner specializing in data engineering, AI, and cloud migration.
Talk to us →How Google Data Commons scaled global dataset throughput 5× with an AI-powered knowledge graph
How a global enterprise turned 30+ siloed datasets into a unified knowledge graph, with AI doing the heavy lifting.
Read story →How SMI Cabinetry turns architectural drawings into structured cabinet estimates in 30 seconds
How SMI Cabinetry automated cabinet estimation — from architectural drawing to structured BOM in 30 seconds.
Read story →How Saudi Aramco cut media response time from 28 days to 12 with an AI-powered engine
How Saudi Aramco’s media team went from 28-day response cycles to 12 days with four AI agents.
Read story →