MUVN connects drivers with available cargo space with people and businesses that need to transport goods. No new fleet, no new logistics. Better use of what's already on the move. Our marketplace grows along corridors, meaning individual routes that become dense enough to be self-sustaining. Whether a route is viable isn't a matter of opinion, but a matter of measurement.

That's why we're building the MUVN Brain: a connected layer above our systems where key performance indicators (KPIs) are defined once and humans and AI agents alike receive the same response. There will always be more than one system. The difference lies in whether they communicate with each other. You're building that.

You'll be our first dedicated data expert, reporting to our CFO and working directly with the founding team, marketing, and operations. You won't inherit a data platform; you'll decide what it becomes. And you'll answer the questions that drive our bottom line: Which corridor will tip next? Where are we losing demand? Which channel is generating matches instead of downloads?

You work AI-First: AI takes care of the execution, you are responsible for intention, logical thinking and correctness.

We are looking for you to join our team full-time starting October 1st, 2026 🚀

Tasks

One truth, not five

You define our key performance indicators (KPIs) across the entire funnel, once and for all, with a clear naming convention and a definition for each metric. Parts of these already exist, but they aren't operationalized anywhere. You reconcile the sources, explain discrepancies, and ensure that every question has exactly one answer, regardless of who asks it.

Making the marketplace measurable

Our unit of analysis is the corridor, not the city or the country. You visualize supply, demand, fill rate, and time-to-match at this level, including cases that don't appear in any standard funnel. A driver who can't find anything is a signal for us, not a missing data point.

Building the MUVN Brain

Today: direct SQL, Python scripts, API pulls. You're transforming this into a lightweight, connected stack and deciding on the architecture and tools, from the data warehouse to the transformation layer and the BI tool. Each tool is only brought in when it justifies its cost, not before. Five reliable models are preferable to fifty semi-familiar ones.