Network VC runs a shelf of thematic funds. Each one needs a different LP.

A platform of funds, not one fund

Most venture firms raise one fund and spend three years deploying it. Network VC works differently. The San Francisco Bay Area firm, founded in 2019 by managing partner Alexander Soroka, runs an ecosystem of angel syndicates and small, thematic funds instead of a single blind pool.

Each vehicle is deliberately narrow. A fund backs a defined pipeline or sector, and individual partners run their own vehicles on shared infrastructure. Network VC calls itself an operating system for independent VC professionals, and it sits inside the wider Startup.Network platform Soroka has run since 2015.

Its screen is specific. At seed, Soroka looks for what he calls the three Ts: team, technology, or traction. At late seed and Series A, a syndicate typically wants to see about $1 million in revenue, or a reputable US venture firm already in the round.

A shelf of thematic funds

Capital is organized by thesis, not by vintage. Network VC runs a fund for Y Combinator companies, a fund for Andreessen Horowitz Speedrun graduates launched in fall 2025, a defense fund backing dual-use and Ukrainian-American startups, and a longevity fund in formation. The portfolio spans roughly 100 companies, most of them US-based, with about a fifth of its civilian startups carrying Ukrainian roots.

The exits are named and dated. Volumetric Biotechnologies, a 3D-bioprinting company Network VC backed in 2020, was acquired by 3D Systems in October 2021 for $45 million upfront plus earnouts. Soroka describes it as a 3x.

The rest of the book spans unrelated sectors. The defense fund alone tracks more than a dozen Ukrainian and Ukrainian-American startups, including Bavovna, which builds navigation for GPS-denied zones, and PANOPTES AI, which automates control of optical-electronic air-defense networks. Alongside them sit US dual-use companies like Sierra Turbines and Orbital Operations. On the civilian side the portfolio runs from Databento in financial market data to robotics startups like Mbodi and Vassar Robotics.

Why many funds change the fundraising math

A firm raising one fund raises from one LP base. A firm running a defense fund, a longevity fund, a Y Combinator fund, and a Speedrun fund at once is running four fundraises. Each has a different natural investor.

The family office comfortable underwriting dual-use defense is rarely the one anchoring a life-extension thesis. The operators who want Y Combinator exposure are a different pool again. Network VC sources across all of it, but every vehicle needs its own base of LPs whose mandate matches the thesis.

That is the practical constraint behind the model. As the fund shelf grows, fundraising stops being about volume and starts being about fit.

Where Altss fits

Each new thematic fund raises the same question: which family offices and independent LPs actually hold a mandate for that specific vehicle, whether defense, longevity, AI, or a named-accelerator cohort.

That is the job Altss does. It filters the allocator universe by sector mandate and family-office profile, then surfaces verified decision-makers. The output is a per-fund outreach list, not one database worked across every vehicle. As each batch fund opens, the list is rebuilt around that fund's thesis.