Outlit wants AI agents to track every business customer

A profile that keeps updating

Founders Josh Earl and Leo Paz say companies want to automate customer work but lack timely information about what their customers are doing. Outlit combines internal records, including call logs and sales data, with external sources such as social media, LinkedIn and employer websites.

The result is a one-page profile that is continually updated with information the system finds about a person. Outlit’s agents organize the signals, draw conclusions and recommend next steps, including whom to contact and what material to send.

Outlit is still pre-launch and mainly works with companies selling software to other businesses. Earl says the startup does not track individual consumers and does not plan to enter that market, arguing that the data would be unreliable.

The product is meant to give sales and customer-success teams a single place to review a customer’s activity, check the sources behind each signal and mark what is relevant. Earl says it may be especially useful when a deal is at risk and staff do not have time to assemble the picture themselves.

That urgency is part of the product’s appeal—and a reason its boundaries matter. A public post can be about a wedding, a new child, illness, divorce or a death in the family. Earl said he had not previously considered that such details could enter the system. Outlit does not seek them out, he said, but it can technically collect anything published online.

What counts as a sales signal

Asked whether people would be notified or given a way to opt out, Earl said no. He argued that Outlit does not break the law and has the right to track signals. The company’s website says customer data is protected and anonymized and is not used to train models.

Earl’s framing is that companies already gather customer information, but people cannot process it fast enough. Automation makes it possible to collect and combine those signals continuously and at greater scale. He expects agents eventually to understand and support human relationships better than people do.

I think the more consequential question is not whether Outlit can assemble a fuller profile, but whether sales teams should treat every publicly available detail as useful context. The system’s ability to distinguish relevant information from irrelevant information is presented as a feature, but the person being profiled has no say in that judgment.

Earl says continuous monitoring will happen regardless of what it is called. That may be a plausible forecast of where sales software is headed; it is not, by itself, a case for making every detail discoverable to the people trying to close a deal. The sharper the profile becomes, the harder it will be to separate better service from surveillance.