Clearspace Wants to Build the Intentionality Layer for the Internet

The modern smartphone is a battleground. On one side, you have your actual intentions—checking a message, looking up directions, calling a friend. On the other, an army of applications fighting for a few more seconds of your attention, using every psychological trick in the book. This friction is the problem Clearspace is tackling. The Y Combinator W23 batch company isn't just building another screen-time tracker; they're trying to build an active filter for your digital life.

Their approach is to create what they call an "intentionality layer." Instead of asking users to rely on willpower alone, Clearspace is developing an agent that sits between you and the internet. This agent processes network traffic in real-time, making decisions based on natural language rules. The goal is to block or filter the content that triggers compulsive behavior before it ever has a chance to hijack your focus.

This isn't a simple content blocker. The core of their technology relies on machine learning models capable of classifying network traffic with nuance. It's a data-heavy problem that requires understanding context, not just keywords. That's where the company's current hiring push comes in. They're looking for a Research Engineer to take ownership of their production model.

The role is focused squarely on the practical challenges of deploying ML in a demanding environment. The job description emphasizes a mindset that goes beyond just tuning algorithms. They're looking for someone who obsesses over the entire pipeline: how to gather more data volume, how to featurize that data intelligently, and how to design models that meet specific inference requirements. It's a signal of the company's engineering culture—moving from research to production requires solving the whole system, not just the model.

Clearspace has gained traction by promising a more deliberate relationship with technology. Their app has been featured in places like the Huberman Lab podcast and the New York Times, indicating a growing appetite for tools that address digital well-being with more than just a simple dashboard. The challenge ahead is scaling their classification engine to handle the sheer variety and volume of network traffic, a problem that sits at the intersection of time-series data, privacy, and user experience. It's a complex technical mountain to climb, but the potential payoff is a genuinely quieter phone.