Most founders pitch a vision before they have a product. Max Ruderman went the other way around. The first version of Harmonic was a scraper pointed at Delaware corporate filings, the dull state paperwork that records the precise moment a company stops being an idea and becomes a legal thing.
If you want to find startups before anyone has heard of them, that filing is the earliest possible signal. Ruderman built a way to read it at scale, then kept reading everything else the web leaks about a young company: headcount, hiring pages, web traffic, who the founders worked with last. Stitch enough of those fragments together and you get something investors had never really had before - a live map of the startup world from the day each company is born.
Today he is the CEO of Harmonic, a New York company that calls itself "the source of truth for startups." The customer list reads like a roll call of the people who decide which companies get funded: Floodgate, Craft Ventures, a16z, Accel, plus operators at Brex, Notion, Vouch and Carta. Several of them bought the product first and then asked to put money into the company that made it.
The injury that started it
Before any of this, Ruderman was a Google engineer who poured his spare hours into jiu-jitsu. He sunk all his time into the mats. Then an injury benched him, and the time he used to spend training had to go somewhere. It went into building a company. He has described the entrepreneurial itch arriving almost by accident, as the thing that filled the gap a sport left behind.
It is a small detail, but it explains a lot about how he operates. Harmonic is a company built on repetition and endurance: refresh the data, refresh it again, keep it current, never let the record go stale. Ruderman did not come to that discipline from a business school case study. He came to it from training for things that take a long time and reward people who keep showing up.
Six and a half years inside Google's data
At Google, Ruderman did a little of everything that touches data: analytics, data warehousing, machine learning and modeling, and straight software engineering. His last role was as a senior software engineer on a Search team, building tools so Google could run user-experience research and design at scale. The throughline is obvious in hindsight - he spent years learning how to make enormous, messy datasets behave, and then went looking for a market where nobody had done it yet.
Private markets were that market. Public companies are documented to exhaustion. Private startups, especially the early ones, are a fog of LinkedIn pages, press releases and word of mouth. The incumbents, Crunchbase and PitchBook, leaned toward later-stage companies and information that aged badly. Ruderman's bet was that the early fog could be cleared with the same engineering he had been doing at Google, just aimed at a different problem.
From COO to the corner office
Harmonic was co-founded by Bryan Casey and Ajay Sohmshetty, with Ruderman joining and running operations as Chief Operating Officer. When Casey and Sohmshetty later left to pursue new projects, Ruderman stepped up to CEO. It is the kind of transition that can wobble a young company, and instead it sharpened one. The product kept its early-stage edge while the company grew into a real organization, recruiting leaders out of LinkedIn, Dropbox and Google.
The money followed the data. A $7 million seed led by Craft Ventures was followed in late 2022 by a $23 million Series A led by Sozo Ventures, with Craft Ventures again participating - roughly $30 million in total. The telling part is who else showed up. Floodgate was a customer before it was an investor. Fin Capital uses the product daily. When the people who buy your software also want to own a slice of your company, the usual gap between sales pitch and reality has mostly closed.
Meeting investors where they already are
Ruderman is careful about what Harmonic is not. It does not try to replace an investor's network. It fills the gap around it - the out-of-network companies a firm would never stumble into through warm introductions alone. The platform plugs into LinkedIn connections, email, calendars and CRM data, then pings a partner when someone in their orbit quietly starts something new. A browser extension can tell a firm, in a second, which of its own people already know a founder it just discovered.
That philosophy - meet people where they work rather than force them onto a new standard - is why Harmonic spread the way it did. Investors did not have to change their habits. They just got a sharper version of the habits they already had.
The AI turn
The latest chapter is an investment agent. Built on LangGraph and LangSmith, it turns plain-English questions into precise startup searches, generates market maps and research reports on its own, and handles tangled criteria - location, funding history, investor connections, who knows whom. The numbers Harmonic reports are blunt: research that used to take hours now takes under a minute, with roughly a 30 percent lift in searches that actually surface something useful.
Ruderman, still an engineer at heart, talks about the tooling the way a builder does, not a marketer: "This UI is invaluable for debugging - instead of rerunning every node, we can directly inspect graph state at any point, make changes, re-run from that point, and observe the difference." It is the same instinct that started the company - get inside the machine, see exactly what it is doing, fix the part that is wrong.
Remote, then not
Harmonic began life as a fully remote team. Ruderman later reversed course, moving to a hybrid model with offices in New York and San Francisco, on the view that in-person time matters for keeping a company aligned. It is a quietly revealing decision from someone whose entire product argues that you can know a great deal about a company from a distance. Even the founder of the source of truth for startups thinks some things still need to happen in the same room.
When he is not running the company, the endurance theme keeps showing up. He trains for the New York Marathon with an in-house group called the Harmonic Running Club, and off the clock he mountain bikes with his wife and dog. He built a machine that never stops refreshing data, then spends his weekends doing things that reward exactly that: keep going, keep going, keep going.
What he is really building
Strip away the product details and Ruderman's ambition is larger than a better search box. He frames Harmonic's job as improving capital efficiency in private markets - getting the right money to the right founders at the right moment, instead of leaving promising companies undiscovered and undercapitalized because they happened to sit outside someone's network. Transparency, in his telling, is not a feature. It is the whole point.
For someone who arrived at startups sideways - by way of a Chinese-and-economics undergraduate detour, a labor relations degree from Cornell, years inside Google's data guts, and a jiu-jitsu injury that freed up his evenings - that is a coherent place to land. Find the signal early. Keep it honest. Let the people who build the future actually find each other. The Delaware filings were just where he started reading.
Profile compiled from public sources including Harmonic's About page, the World of DaaS podcast, TechCrunch, LangChain's customer case study, Signature Block and The Org. Funding and customer figures reflect publicly reported numbers and may have changed.