Welch's email arrived on a Tuesday, eleven days after the Arcline incorporation.
It came to Marcus's public business address — the one on the Threadline website, which was the address used by leads, journalists, and conference organizers. The email was professionally calibrated: not overeager, not understated. It described Arcline Data Partners as a strategic advisory firm focused on data intelligence applications in institutional investment. It said Welch had been following Threadline's work with interest. It proposed a thirty-minute introductory call.
Marcus read it twice. It was well-written. If he had not spent the previous two weeks building a dossier on Thomas Welch and a detailed understanding of the entity behind him, he would have taken it at face value.
He forwarded it to Elaine and to Marsh simultaneously.
Elaine called within the hour. "Accept the call. Propose next Thursday — gives us time to prepare."
Marsh called forty minutes later. "I've reviewed the Arcline incorporation. The entity is clean on its face. If Welch makes any proposal in the call, do not engage with specifics until I've reviewed the proposal. Acknowledge interest and request materials."
"Understood," Marcus told both of them.
He composed his reply to Welch: *Thanks for reaching out. Happy to connect — Thursday works. Here's a scheduling link.*
Welch confirmed within twenty minutes. 10 AM Thursday.
---
The briefing with Elaine was Wednesday afternoon, over a call she ran from what sounded like a car. She laid out three objectives for the Thursday call: establish what Arcline was actually offering, understand what intelligence about Threadline they were operating from, and listen for any indication of the principal network's current priorities.
"Don't press," she said. "Let him lead. The more he talks, the more you learn."
"And if he makes a direct proposal?"
"Express interest without commitment. 'That's interesting, let me look at this more carefully' is a complete sentence."
Marcus thought about Welch — the Wharton degree, the Eastbridge papers, the conference panel. A professional who had spent years sitting across from founders and executives and presenting complex propositions in ways designed to look simpler than they were.
"He's going to be good," Marcus said.
"Yes," Elaine said. "He usually is."
---
Thursday 10 AM. Welch appeared on screen in a home office that was professionally appointed without being ostentatious — the kind of background that communicated substance over display. Late forties, silver at the temples, a quality of unhurried confidence that Marcus recognized as the result of years of practice rather than natural ease.
"Marcus. Thank you for the time."
"Of course. Tell me about Arcline."
Welch smiled slightly — the smile of someone who appreciated directness. "We work at the intersection of data intelligence and institutional capital allocation. Our clients are long-term oriented — sovereign funds, large family offices, endowments. They're trying to make decisions about complex systems over multi-year horizons, and the data tools that exist in the market are either too tactical or too backward-looking." He paused. "What you've built at Threadline is neither of those things."
"What is it, in your assessment?" Marcus said.
"It's a structural intelligence tool. It doesn't just report what happened — it models the relationships that produce outcomes. That's architecturally different from what most alternative data providers are doing." He looked at Marcus. "Your approach to entity disambiguation is particularly interesting to our clients. The ability to surface hidden relationships at scale — that's not a feature, it's a capability that changes the analytical frame."
Marcus listened to this carefully. Welch's description of the Threadline methodology was accurate — more accurate than the deliberately-understated IP paper had been. He was presenting a different picture to Marcus than the one that had been published. Which meant either the paper had been written for a different audience, or the paper had been designed to lower Marcus's guard while the real assessment was this.
"What kind of engagement are you proposing?" Marcus said.
"At this stage, I'm not proposing anything specific. I'd like to understand your roadmap — where Threadline is going in the next twelve to eighteen months — and explore whether there's a complementary relationship. Our clients generate data. You have the architecture to extract signal from it. That seems like a natural fit."
"What kind of data do your clients generate?"
"Transaction data, primarily. Cross-border capital flows, asset acquisitions, institutional positioning." A pause. "Proprietary data that isn't publicly available. Significant volume."
Marcus thought about what Welch had just described. Cross-border capital flows. Institutional positioning. The precise categories that the Depth project's classified layer was already analyzing through the working group's access.
He thought about the possibility that Welch was offering him access to data that would, if he ingested it, give the principal network visibility into what Threadline's methodology could surface from their own financial activity.
A data trojan. Not a technical intrusion — a business one. Offer the data, watch what the analysis finds, understand what the system can see.
"What's the commercial model you're imagining?" Marcus said. "Hypothetically."
"Data licensing on our side, analytical output on yours. Revenue share on any strategies our clients implement using Threadline's signals." Welch's expression was open, relaxed. "We'd want some degree of exclusivity on the specific client data — our clients are sensitive about their information being analyzed for other parties' benefit."
*Exclusivity on the client data.* If Marcus agreed to that, he would be analyzing their data in a silo, with no ability to cross-reference it against other sources or share findings with, for example, a government working group.
The structure was elegant. It would have been very compelling to a founder who didn't know what he was looking at.
"Interesting," Marcus said. "Let me think about this and talk to my team. Can you send over a one-pager on Arcline's current client structure and data access capabilities? I'd want to understand the data volume and coverage before we go further."
Welch nodded. "I'll have something to you by end of week."
"Great. I'll be in touch."
The call ended.
Marcus sat for a moment. He pulled up a blank document and wrote, from memory, every specific claim Welch had made and every piece of information Welch had demonstrated knowledge of. He read the list back.
Welch had known about the disambiguation architecture. He had known that Threadline was moving toward supply chain intelligence. He had specifically framed the proposal around cross-border capital flows — which was not a logical fit for Threadline's public product profile, but was a logical fit for someone who knew about the Depth project.
He called Elaine.
"He knows about Phase Two," Marcus said.
A silence that was different from her previous silences. Longer. More careful.
"Tell me exactly what he said," she said.
He told her. When he finished, she was quiet for four seconds.
"Cross-border capital flows," she said. "That's not a deduction from your public profile. That's a specific."
"Yes."
"He has a source," she said. "Inside the working group or adjacent to it."
"Or he has a source who read the oversight assessment."
Another silence. "I'll escalate this today." A pause. "Marcus. The proposal he described — the data exclusivity structure—"
"I know what it was."
"Good." A pause. "Don't respond to Welch until I come back to you. This changes things."