Eagle's 85/5 Pitch: Lightspeed's MEP Roll-Up Is a Two-Sided Sourcing Event
Eagle's June 2026 HN post is a live playbook for the AI services roll-up. Here's where to find the engineers it will shake loose, and the FDEs it can't.
Eagle's June 2026 "Who is hiring?" post on Hacker News reads like a normal NYC startup ad: $150K to $300K, onsite, Founding Member of Technical Staff. It isn't. It is the cleanest public artifact yet of the AI services roll-up, and if you source technical talent for a living, it is a two-sided event you should already be working.
On one side, Eagle is hunting a very narrow pool of Forward Deployed Engineers, Staff SWEs, and ML Engineers to embed inside firms it acquires. On the other, every firm it buys becomes a pool of licensed civil, structural, and MEP engineers under new management. Both sides are reachable. Most recruiters are only watching one of them, and it's the wrong one.
What Eagle actually said, and what it means
Eagle is "backed by Lightspeed Venture Partners" and "acquires and transforms civil, structural, and MEP engineering firms with applied AI." The thesis they're recruiting against is a single sentence repeated across their JDs: 85% of what engineers do today is theoretically automatable, yet less than 5% has actually been touched by AI. They call it the largest gap of any profession.
The open roles, all NYC onsite, all in the $150K to $300K band: Forward Deployed Engineer, Staff SWE, ML Engineer, Applied AI Engineer, Design Engineer, Member of Technical Staff. The FDE spec is explicit. You will "embed inside the firms we acquire, sitting beside structural, civil, and MEP engineers who've often never worked with a software company" and "wire up integrations to legacy Windows-native AEC software." That is not a marketing flourish. That is the operating model.
The founding team backs it up. Mayank (ex-Bloomberg, ex-SEC, ex-Ethic, son of a civil engineering firm owner) plus a CTO with 15+ years across Arup, Palantir, Pivotal, and Replica. The Arup-Palantir combination is the tell. This is a Palantir-style FDE motion applied to AEC.
The category, in one paragraph
CNBC named this category on June 8, 2026, calling it the "AI rollup," with the $7.6B Janus Henderson (General Catalyst/Trian) and $6.3B GBT (Long Lake) take-privates as the latest public-market signals. Newcomer broke the Lightspeed thread in December 2025: partners Isaac Kim and Amish Desai are running roll-up plays in engineering services and healthcare. General Catalyst has co-created at least 10 of these companies. The names worth memorizing: Eudia (legal), Titan (IT/MSPs), Long Lake (HOAs, $670M raised in under 2 years), Accrual (accounting), Crescendo (BPO), Dwelly (UK lettings). Thrive Holdings has over $1B earmarked. Crete Professionals Alliance, Thrive-backed, plans $500M+ on accounting acquisitions using OpenAI tooling.
This is no longer one fund's experiment. It is a vehicle class. And the labor-market consequence of that vehicle class is the part nobody is sourcing against yet.
Side one: the FDE pond is tiny, and everyone is fishing it
"Forward Deployed Engineer" as a US title returns roughly 2,100 profiles. The top feeder is Palantir, by a wide margin. The rest is Scale AI, Sourcegraph, Cursor, Modal Labs, and Cresta. That is the entire addressable pool for AI services roll-up hiring at the FDE level. Eagle is competing with Eudia, Titan, Crescendo, Accrual, and a handful of stealth Sequence Holdings types for the same 2,000 people.
Three things follow from that.
First, title search is already broken here. The people who do this work at Palantir don't all carry "Forward Deployed Engineer" on LinkedIn. Some are "Deployment Strategist," some are "Forward Deployed Software Engineer," some are just "Engineer" with a customer name in the description. If you're running boolean strings on title alone, you're missing half the pond. This is exactly the friction we built Refolk to remove: you describe the person in plain English, including the work they actually did, and get a ranked shortlist across GitHub, LinkedIn, and the open web rather than a title-matched list that excludes the people you most want.
Second, comp will inflate fast. Eagle's $150K to $300K band is below FAANG ML and well below frontier-lab ranges. They are trading cash for equity narrative and domain access. That works for the first ten hires. It does not work for the hundredth, and it does not survive a Cursor counter-offer. Recruiters at hyperscalers and frontier labs should be tracking Eagle FDEs on a 12-month timer. If integration drags, those people will move.
Third, ex-Palantir FDEs are the new ML researchers of 2021. Every roll-up wants them. Every roll-up is writing nearly identical JDs. Differentiation now comes from how you reach them, not what you offer.
Ex-Palantir FDEs are the new ML researchers of 2021. Every roll-up wants them. Most are writing the same JD.