Everyone is building something with AI right now. Some use chatbots, others automate tools like Slack or HubSpot. On the outside it looks slick, but under the hood it often becomes a mess. AI workflows only work if the data behind them is right. And that’s where we come in.
At i-spark, we have been building Composable Data Hubs for years. These are modular data platforms that ingest, transform and activate structured data. They power dashboards and deliver enriched data feeds to systems such as email tools, advertising platforms and now AI agents. So when we talk about AI, we’re not changing direction. We’re building on what’s already there.
AI Workflows × Composable data
Roughly speaking, there are three types of AI-related projects we encounter.
First, there is simple automation: tools like Make or n8n linked to business apps. You input some logic, call an LLM and get an email or Slack message as output. These workflows are widely used, and if that’s your use case, we can obviously help you.
Second is the data infrastructure. This is what we have always done. A composable data hub that delivers both insights (dashboards, reports, analytics) and enriched data feeds as structured output, ready for activation. These feeds used to flow to marketing tools. Increasingly, they now feed AI-driven decisions.
Where it gets interesting is when these two worlds meet: AI workflows built on top of strong, structured data ecosystems. That’s our focus: Not just writing prompts or developing connections, but creating value by applying ecosystems that give AI the clarity and context it needs to be truly useful.
One example: when the model looks smart but is wrong
In one of our MVPs, we used n8n to generate daily updates on product data, which were dropped directly into Slack. The idea was to replace manual analysis with something faster, fed by both internal data and external context such as competitor promotions or market news.
The query generated by LLM tried to add up the number of daily users to compare this week to last week. But these counts were snapshots, not flows, so adding them together made no sense. A junior analyst might have made the same mistake. A senior would probably not. That’s the point.
It’s not just about fixing syntax. It’s about understanding what the data means and how to detect faulty logic that looks right. That’s why we believe data professionals have a key role in building AI workflows that leverage real business data.
“We’ve spent years building structured, governed data systems. Now we use that foundation to drive AI workflows that are not only impressive, but reliable. “
Why data teams are important in AI workflows
Building AI tools is a team effort. Front-end developers and automation builders all play their own roles, especially in creating great interfaces and flows. But when your automation interacts with structured, sensitive, mission-critical data, you need people who know how that data is modeled, calculated and managed. That’s where we – seasoned data professionals – really add value, by making sure the AI has clean inputs and applies the right logic:
- We know which numbers can add up and which cannot.
- We think in terms of lineage, context and consistency.
- We see it when AI output does not make sense, even if it looks fine at first glance.
- We build with governance, access control and sensitivity in mind.
It’s the difference between making a tool that works and a system that lasts. Making AI reliable, not just impressive.
Or as we say:
If you take AI seriously, you have to have your data right. And if you take data seriously, you can no longer ignore AI.
What makes a company ready for AI?
Not all AI runs on structured data. But when it does, when it drives decisions, automation or personalization, the basics matter. In our view, there are five building blocks that make an organization AI-ready:
- Clear metadata & semantics: context around what the data means, how it is defined and how it should be used.
- Modular architecture and scalable infrastructure: modular systems that connect and communicate well with each other.
- Trusted data quality & traceable lineage: trusted pipelines with validation, freshness and clear traceability.
- Robust governance and managed access: knowing who can use what, under what conditions.
- AI-ready integration: workflows, promptable layers or vector access points that allow AI to make sense of the data.
These are not just technical check boxes. They allow both teams and AI systems to make decisions with confidence.
Where we stand
We didn’t dive headlong into it when AI first emerged. Not because we weren’t interested, but because we focused on what would be useful, not just new. We researched, experimented, built MVPs. Now that the initial dust has settled and the path to real opportunities is becoming clearer, we are ready to take action.
If you’re building a custom front-end around AI prompts, we’re probably not your team. But if the strength of your AI depends on structured, reliable data, that’s exactly where we come in. So if you’re working on serious AI and you want it to hold up over time, we’d be happy to help.
We don’t change direction.
We are building forward.