Why an AI-ready data platform came first
For OMHU, an AI-ready data platform isn't just about reporting; it's the foundation for how the company plans to work with AI.
OMHU wants to build a data- and AI-centered company. This requires reliable data. AI agents, forecasting, business intelligence, and operational automation all depend on one thing: trusted data.
That's where Weld came in.
Weld is helping OMHU build a single source of truth that enables business users to find insights and run analyses.
Weld is just the first building block in our AI vision. We need to ingest the data, we need a single source of truth, we need metrics defined, and in order to do that, we need a data platform.
– Laurenz Brockmeyer, Director of Data & Analytics, OMHU
Before Weld:
- Multiple BI tools created differing versions of truth.
- Key operational data was unavailable for analytics tasks.
- AI agents in Slack were missing metrics, views, and context they could use to deliver use cases.
About OMHU
OMHU is a Danish design company built on a friendship of more than 30 years. Jonas and Frederik started out trading curated vintage furniture as Kram & Klenodier, and in the spring of 2022 they designed the TEDDY sofa, the modular piece that became the company's signature.
Everything OMHU sells is made to order rather than held in stock, produced in the company's own facility in Poland, and the range runs to 21 colours.
Building the right data foundation
When Laurenz Brockmeyer joined OMHU, the company had ambitious plans for its data platform, and at that time, no centralized data warehouse.
Instead, teams relied on several BI tools, each providing different views of the business.
There were several sources of truth. For example, Total Sales can be defined in various ways.
– Laurenz Brockmeyer
At the same time, many operational data points simply weren't available. While marketing and sales reporting existed, important metrics like shipping costs, payment provider fees, and other operational drivers were missing.
It was only a fraction of our data that was reflected in these SaaS tools. We were lacking very crucial operational data.
– Laurenz Brockmeyer
The team needed a modern data platform that would support long-term growth without creating unnecessary engineering overhead.
Finding the right balance between buying and building
As OMHU designed its future data architecture, one question kept coming up: should they build everything themselves or rely entirely on external platforms?
Rather than choosing either extreme, they opted for a modular architecture built around tools like BigQuery, dbt, Cube, and Weld for data ingestion.
We found a good sweet spot of buy versus build.
– Laurenz Brockmeyer
This approach gives the team ownership of its data platform while removing the burden of maintaining data connectors internally.
For Laurenz, that's a significant advantage.
I'm happy that when Shopify changes the API, it's not on my plate to figure it out. Weld is doing it for us.
– Laurenz Brockmeyer
Powering AI agents with trusted data
Instead of maintaining data pipelines, the team can focus on extending the data platform, building AI agents in Slack for business users, and delivering high-impact work centered on business intelligence, forecasting, and automation.
With reliable data flowing into a centralized warehouse, OMHU is now building the next layer of its platform.
The team is using centralized data for:
- Daily sales and marketing reporting
- Finance reconciliation and invoice validation
- Traffic and conversion analysis
- Geographic sales insights
- AI-powered Slack assistants
Looking ahead, the focus is on expanding AI adoption across the business.
Q3 and Q4 will really be focused on onboarding users and onboarding agents.
– Laurenz Brockmeyer
By creating a trusted semantic layer, OMHU is making it easier for both people and AI systems to answer business questions with confidence.
More than an ingestion platform
Throughout the implementation, Weld became more than a data ingestion tool.
Beyond keeping pipelines running reliably, the team also collaborated on data models, transformations, and metrics as the platform evolved.
Besides your standardized ingestion pipelines, you're helping us build our transformations, our models, and our metrics.
– Laurenz Brockmeyer
Laurenz also highlights Weld's responsive support as one of the biggest differentiators.
Support via Slack works great. That's really making a big difference.
– Laurenz Brockmeyer
People are super nice, super responsive. If an API fails, you fix it.
– Laurenz Brockmeyer
Looking ahead
For OMHU, building an AI-ready company starts with building an AI-ready data platform.
By combining a modular modern data stack with reliable ingestion from Weld, the team now has a single source of truth that can support both today's reporting needs and tomorrow's AI initiatives.
A very good experience, and I think a good European alternative to Fivetran.
Interviewed
Laurenz Brockmeyer, Director of Data & Analytics, OMHU
