Raylu

Somewhere in a private equity office right now, an analyst is building a market map by hand. Tabs open, coffee cold, a spreadsheet growing one company at a time. Raylu was built to make that scene obsolete.

A small team aimed at a very old problem

Raylu is a New York company of about fourteen people doing something the private markets have wanted for decades and never quite trusted a machine to do: find the right companies to invest in. Not a list. The right ones. Venture, growth, buyout, and corporate development teams point Raylu at a thesis and get back a custom market map, enriched intelligence on every target, and a verified path to the founder's inbox. The work that used to eat an analyst's week happens before the coffee gets cold.

Forty-five-plus funds use it. They represent more than $500 billion in assets under management. That is a lot of conviction for a company that, by headcount, could fit around two dinner tables.

What they are now is easy to state and hard to build: an AI sourcing engine that investment professionals actually keep open. Not a demo they tried once. A tool that sits in the workflow, next to the CRM, doing the part of the job that looks like research and feels like data entry. The bet is that the most valuable thing a fund owns is attention, and that attention is wasted the moment a partner spends it assembling a list instead of judging one.

Sourcing was a numbers game played by hand

Deal sourcing has always been a grind dressed up as instinct. Firms pay handsomely for databases, then pay analysts to read around the gaps in them. The good companies - the niche disruptors, the founders not yet on anyone's radar - are precisely the ones legacy databases miss, because being missable is what makes them early.

The irony writes itself. The entire industry exists to find what others overlook, using tools designed to surface what everyone already sees. Raylu's founders looked at that and decided the bottleneck was not effort. It was the map.

Two engineers, one investor, one beer

Raylu came from three friends. Ali Dastjerdi spent his early career at Insight Partners, on the buying side of the exact problem he now sells against. Nathan Ondracek and Sam Ilkka came from Amazon Web Services, where building reliable systems at scale is the whole job. One had watched sourcing break from the inside. Two knew how to engineer their way out of it.

They also made a pact, the kind that sounds like a joke until you realize it is a real one: if the company collapses, they drink a single beer together and move on. It is a low bar for failure and, judging by the funding, an unlikely one to test.

"Deal Engineering," and what it actually means

Raylu calls its approach Deal Engineering: the discipline of turning non-deterministic AI into repeatable sourcing outcomes. Translated out of the pitch deck, it means the system does not just answer a question once. It runs the same messy work - mapping a market, profiling companies, finding the right human to email - reliably enough to bet a fund's pipeline on.

Feed it a thesis and it returns a market landscape with roughly 2.5x more actionable targets than a traditional pull, drawn from a database the company puts north of 57 million companies. Each target arrives with about 72 AI-generated data points and citations, so a partner can check the work instead of trusting it blindly. Then it finds the decision-maker - the company claims a 95% CEO email match rate - and the outreach that follows lands roughly 30% more replies.

The sequence matters more than any single number. Thesis becomes market map. Market map becomes a stack of researched targets. Targets become verified contacts. Contacts become meetings. Each step in that chain used to be a separate tool, a separate analyst, and a separate week. Raylu's argument is that the handoffs between those steps were where most of the time, and most of the missed companies, quietly disappeared. Collapse the handoffs and you do not just go faster. You see more.

There is a quieter feature buried in all of this, and it is the one finance cares about most: the research is auditable. An agentic chat will dig, by the company's account, far deeper than a consumer chatbot, but it shows its work. In a business where being confidently wrong about a company can cost a fund real money, citations are not a nicety. They are the difference between a tool a partner will defend in an investment committee and a toy they will quietly stop using.

Figures per Raylu's own materials - the kind of numbers that look great on a slide and, refreshingly, come with citations attached.

When your lead investor is also your customer

The cleanest endorsement in Raylu's story is structural. HighlandX, which led the $8 million Series A, was reportedly a user of the product before it was an owner of the company. There is no warmer reference than an investor who tried to find reasons not to buy the software and ended up buying the business.

Behind that sits the rest of the cap table - Conversion Capital, Unusual Ventures, and a founders' club of angels who have built things that scaled: Dropbox, Algorithmia, DataRobot, Pylon. Raylu plugs into the tools these firms already run on, syncing with DealCloud, Affinity, and Salesforce, and wraps the whole thing in SOC 2, zero data retention, and end-to-end encryption, because finance does not adopt anything it cannot first lock down.

Make the analyst's grind a setting, not a job

Raylu's stated aim is to make AI-native origination the default for private markets - to take the manual, labor-intensive parts of finding deals and turn them into something repeatable. The point is not to replace judgment. Judgment is the part the partners are paid for. The point is to stop spending it on tab-switching.

It is a strangely humble mission for an AI company in 2026. There is no claim that the software picks winners or replaces the investor's gut. The pitch is narrower and, for that reason, more believable: do the boring 80% so the human can spend all of their judgment on the interesting 20%. Wall Street has heard a thousand promises about disruption. This one is mostly a promise about giving people their afternoons back.

The map gets redrawn

If Raylu is right, the edge in private markets shifts. It stops belonging to whoever can afford the most analysts reading the most documents, and starts belonging to whoever asks the sharpest question and trusts the system to go answer it. That is a different game, and a fairer one for the small firm with a good thesis and no army.

Back to that office, then. The analyst is still there. But the spreadsheet built itself overnight, the founder meetings are already on the calendar, and the cold coffee is just coffee now - not the price of doing the work. That is the scene Raylu is selling. So far, $500 billion in assets under management has decided to buy it.