Kevin Otte joined fal as its first and only data hire when the company had ~30 people. Six months later, they scaled to 80 and suddenly every function needed data at the same time: Finance needed board-ready reporting, Sales needed playbooks, and Product needed analytics coverage for a steady stream of new instrumentation.

This is how fal built a high-leverage analytics function with one person, strong engineering partners, and agent-assisted delivery.

The problem: One person, 80 stakeholders

Kevin’s job was to build the data function from scratch, and keep it running as the company scales around him.

fal had been relying on a notebook-style analytics workflow, which worked well for ad-hoc analysis but wasn’t designed to be the self-serve BI layer a scaling GTM team needs.

“As we started building out our GTM function,” Kevin explained, “stakeholders needed an easy way to navigate the data. Our VP of Sales wanted to build playbooks, a view you open every morning to see what you have in terms of work. The reality was that wasn’t going to be possible with our existing setup.”

What fal needed was a BI tool that felt familiar to business users, and gave Kevin full control over the data model. He looked at a lot of tools, and eventually settled on Lightdash. 

"I spent five minutes in Lightdash and I was like, oh, that was extremely easy. It's quick to build, and the interface makes sense to business users. It checked all the boxes."

But switching tools didn't solve the core problem: one data person, fielding requests from 80 colleagues. 

The solution: Agentic BI

When Kevin saw Lightdash's Agentic BI workflow, he set it up immediately.

Now, a ticket comes in through Linear or Slack, a Cursor agent picks it up, accesses the Lightdash MCP server, reads the semantic layer, builds the chart or dashboard, and creates a preview link that Kevin can review before pushing to production.

The upshot? Kevin can be writing a strategy doc, running a cross-functional meeting, or building out the next part of the data model. All while dashboards are being built in the background.

“I built two complete dashboards from scratch. I described the charts, the filters, the layout and agents handled the repetitive build steps. What used to take a full day took about an hour.”

From answering questions to shipping systems

The speed with which he can ship analytics and cater to requests has been game-changing. 

"And now if someone wants to add another view or another chart, it's as simple as: create a ticket, let the agent build it, review it, done. It takes five minutes."

The bigger unlock is the type of work Kevin no longer has to do. The context-switching, the small, tedious, yet somehow time-consuming requests that used to pull him away from actual data work.

"Imagine before: you have six dashboards, ten charts each. Someone asks you to switch from a general calendar to a fiscal calendar,” Kevin explained. “Now I need to add a fiscal dimension for everything and update all those charts. That would take me at least a day, realistically. Now I just say: create new dimensions for the fiscal calendar, update all these dashboards to reference that. That's a huge time saver."

AI agents for business users 

Kevin also set up Lightdash's built-in AI agents: one focused on revenue, and one for sales ops. Now, when a colleague asks him a question about the data, his default answer is: have you tried asking the agent?

The habit is still forming across the team, but the direction is clear. 

What "agentic" actually means for a team of one

Kevin's situation is a useful stress test for what Agentic BI means in practice, showing what data teams can do when they're not spending their days clicking through UIs and processing ticket queues.

At fal, one data person alone is supporting exec reporting, GTM dashboards, product analytics, sales playbooks, and a long tail of ad hoc requests for a company that tripled in size in six months. 

He's doing it with Lightdash, dbt, and a workflow where AI agents handle the grunt work while he focuses on the architecture, the governance, and the stuff that actually requires a human to think.

"It's like having data agents do the repetitive work that shouldn’t require senior attention, so I can focus on higher-leverage tasks."

Want to see what Agentic BI looks like for your team? 

See how Lightdash can help your business users find insights without bottlenecking your data team. Start your free trial or learn more about Agentic BI.