How Justin Ernest Built a $400M Startup Portfolio Without a Traditional VC Fund

Justin Ernest spent over five years at Playground Global investing in deep tech and leading fundraising efforts before striking out on his own. Rather than endure the 12-to-18-month process of raising a formal venture fund, he built Sabertooth VC on a radically different model. According to TechCrunch, Ernest identified that family offices and smaller institutional investors were hungry for exposure to the fastest-growing AI companies but found themselves locked out of those cap tables.

His solution was to leverage his dual network of founders and investors to secure allocations in high-profile, later-stage companies. He then packaged these as individual deals offered to roughly 30 smaller institutional investors through special purpose vehicles, or SPVs, which function as single-deal funds. This structure allowed him to move faster than traditional funds and avoid the overhead of a permanent capital vehicle.

Special Purpose Vehicles have long been used by venture capitalists to carve out specific deals for select backers. Ernest's innovation was to make them the core of his entire investment strategy rather than an occasional side channel. The SPV model lets each investor participate deal-by-deal without committing to a multi-year fund structure, giving them more control over where their capital goes.

For the companies receiving investment, the structure is largely transparent. They get capital from a single entity, Sabertooth, even though the money is ultimately pooled from dozens of separate limited partners. This has allowed Ernest to write checks ranging from $10 million to $100 million, according to Sabertooth's own materials, without needing to manage hundreds of millions_fm_ millions in committed capital upfront.

Ernest's investments have landed in some of the most competitive names in technology. TechCrunch reports allocations in Anthropic, the AI safety and research company; Anduril, the defense technology contractor; and SpaceX, Elon Musk's aerospace manufacturer. These are not companies that accept capital from any source. Their inclusion in his portfolio signals that Ernest's network extends to the very top tier of founders and boards.

The concentration is deliberate. Sabertooth's website states the firm pursues a "concentrated investment strategy" focused on "the most ambitious entrepreneurs solving the world's most critical problems." This approach mirrors the philosophy of Ernest's former employer, Playground Global, which also focused on deep technology rather than consumer apps or software-as-a-service plays.

The traditional venture fund model has a well-documented access problem. New managers can spend over a year raising their first fund, only to find that the best deals have already been captured by established firms with decades of track records. Limited partners, meanwhile, increasingly want direct exposure to specific companies rather than blind pools of capital managed by general partners.

Ernest's structure addresses both pain points simultaneously. He gets to market faster by avoiding the fund-raising slog, and his investors get transparency into exactly which companies they own. This is particularly valuable in the current environment where AI valuations have compressed and LPs are more cautious about committing to long-dated fund structures. The model is not without precedent, but its scale, nearly $400 million deployed in 12 months, suggests it has moved from niche experiment to viable alternative.

The SPV-heavy model carries distinct risks that traditional funds do not. Each vehicle requires separate legal documentation, tax treatment, and investor relations. The administrative burden scales linearly with the number of deals, rather than being amortized across a single fund structure. For Ernest, this means building operational infrastructure that most seed-stage firms never need.

There is also the question of follow-on capital. Traditional funds reserve capital for pro-rata rights and subsequent rounds. SPV-based investors must either raise new vehicles for each follow-on or risk dilution. Given that Anthropic, Anduril, and SpaceX are all companies that raise capital frequently and at enormous scale, maintaining ownership percentages will require continuous fundraising from Ernest's LP base. Whether that base remains as enthusiastic in a downturn is an open question.

The emergence of Sabertooth at this scale suggests the venture industry is fragmenting into specialized structures that match specific investor demands. The uniform fund model, 10-year lockups, 2-and-20 fees, and quarterly capital calls, is no longer the only viable path. For certain profiles of LP, particularly family offices with specific sector interests, deal-by-deal access is worth the additional complexity.

Ernest's background at Playground Global, followed by founding NextGen Venture Partners and stints at Sidewalk Labs and Advent International, gave him the exact combination of founder relationships and investor credibility needed to pull this off. The model is not easily replicated by newcomers without equivalent networks. Still, its success will likely inspire imitation, and that could reshape how capital flows to the most competitive private companies in AI and defense technology.