The third Monitor story broke things open.
It was larger than the first two — a thirty-five-hundred-word investigation into a network of school construction contracts across four states, involving fifteen vendors and approximately $180 million in public funds. The data visualization Threadline produced for it was the most complex the tool had generated for publication: a three-tiered network graph with animated relationship highlighting that the Monitor's graphics team had spent two weeks turning into something a general audience could follow.
The story ran on a Sunday morning and within six hours it had been read by enough people that Marcus's name appeared in a sentence in a major national newspaper for the first time. Not prominently — he was mentioned as the founder of Threadline in a paragraph about investigative data tools — but there.
He read the paragraph on his phone. Noted it. Put the phone down.
The following morning, his inbox had twenty-two new messages. Most of them were follow-up inquiries from journalists or compliance professionals who had found his email address through the company website. Three of them were from people he didn't know, with addresses that had the particular structure of professional aliases rather than real names.
He forwarded the three to Marsh.
She called him that evening. "The first one is a PR firm fishing for a quote. Ignore it. The second appears to be a law firm — almost certainly the same firm connected to the first Monitor story. They haven't made any specific assertion yet, but the email is designed to be threatening without being specifically actionable."
"What do you recommend?"
"No response. We've documented everything. If they want to file something, they file something, and we respond to the filing. Don't give them a pre-filing conversation."
"And the third?"
A brief pause. "The third is interesting. It's from an individual named Aaron Reiss. He's a partner at a hedge fund called Meridian Capital Group."
Marcus stopped. "Not Meridian Partners."
"No. Different firm, different sector. Meridian Capital Group is a quant fund. Moderate size, good reputation, focused on alternative data strategies."
"What does he want?"
"His email is four sentences. He says he's been watching Threadline's output with professional interest, he has thoughts about an application of your methodology that he'd like to discuss, and he's happy to sign an NDA before the conversation."
Marcus thought about this. A quant fund interested in procurement anomaly data was not an immediately obvious connection. Procurement patterns could be relevant to equity research — contract awards affected company revenues — but it wasn't the most direct path to an alpha strategy.
"Set up a call," Marcus said. "After I've done a background check on Reiss."
Reiss checked out cleanly. Partner at the fund for six years, quantitative background, public papers on alternative data strategies. The fund's public track record was good. Nothing alarming.
The call was brief and unusual.
Reiss was direct in the way that people from quantitative finance tended to be direct — not aggressive, just operating on a preference for precision over social lubrication. He wanted to know one thing: whether Threadline's anomaly detection could be applied to private-sector procurement data.
"Most large corporations have internal procurement databases," he said. "The data isn't public, but it's sometimes obtainable through data licensing agreements with vendors who handle the transactions. If your methodology could identify anomalies in corporate procurement — not just government — the applications for equity research are significant."
Marcus listened to this. He thought about the architecture. "It's possible in principle. The normalization layer would need to handle private data formats, which are less standardized than government formats. But the relationship graph approach is source-agnostic."
"Would you be interested in a research collaboration? We fund the data licensing, you apply the methodology, we share any findings."
"What's the commercial arrangement?"
"Revenue share on any strategy we develop using the outputs. I'd propose fifteen percent of net revenue from affected strategies, capped at a defined amount."
Marcus thought about the timeline. He thought about Depth, and the Series A, and the existing customer pipeline. He thought about whether adding another track was possible or whether it was the kind of thing that looked possible from the outside and killed companies from the inside.
"Not now," he said. "Six months from now, possibly. I'll come back to you."
"Understood." No pushback. He filed a note. "If your public-data product reaches the right scale, the conversation changes anyway. We'll talk."
The call ended.
Marcus sat with it for a moment. Then he opened the System display and looked at it — not because it had updated, but because he wanted to think about the three tracks now running in parallel: Threadline's commercial growth, the Depth project, and the emerging possibility of a financial intelligence application that he had not originally planned for.
*Build what others cannot follow.*
He was starting to understand, more fully, that this was not a single thing. It was an expanding territory.