BAG Ventures Closes $11.3M Fund I for Enterprise AI Startups

A small fund with a pointed thesis

BAG Ventures, a firm started by two Google alumni, has closed an $11.3 million debut fund aimed at early-stage artificial intelligence startups.12 It is a modest amount by the standards of recent AI dealmaking. The fund's thesis is the notable part. The founders are betting that companies are finishing their period of AI trial runs and will soon pay only for products that clearly earn their keep.2

The firm is led by Bonita Stewart, a former Google vice president, and Jackson Georges Jr., formerly a partner at CapitalG.2 Fund I is set up to invest in "all things AI."1 In practice, that means a defined set of sectors: AI infrastructure, compute, physical and edge AI, security, governance, and vertical SaaS.2

Investing before the close

BAG Ventures was not waiting on the final close to start writing checks. The firm spent roughly two years investing from the fund as capital came in, and it has already built a portfolio of 10 companies.2 Three of them have been named:

• SXD, a software company
• BizTrip, an AI travel agent
• Nomadic, an agentic reasoning platform2

Checks run from $100,000 to $500,000. The team plans to deploy the rest of the fund over roughly the next two years.2

Those figures put BAG Ventures firmly at the pre-seed and seed level. The fund is small and has already been partly deployed across 10 companies. At these check sizes, the remaining capital would likely support another batch of companies rather than large follow-on rounds. The exact number will depend on how much the firm holds back in reserves, which it has not detailed.

Why the "will actually pay for" framing matters

The central claim is that enterprise buyers are becoming more selective.2 Over the past few years, many large organizations have run pilots, set aside innovation budgets, and tested generative AI tools without committing to long-term contracts. BAG Ventures is betting that this pattern is ending. In its view, the startups that succeed will be the ones that can show measurable value to customers that are scrutinizing spend more closely.2

The sector list fits that view. Security and governance are categories companies tend to fund out of necessity, especially once AI systems handle sensitive data or make consequential decisions. Vertical SaaS sells into specific industries where return on investment can be measured in narrower terms. Infrastructure, compute, and edge AI sit lower in the stack. Their buyers typically pay for capacity and performance rather than novelty.2

Together, these categories suggest a firm leaning away from consumer-facing AI apps and broad horizontal tools toward the less visible plumbing and compliance layers that enterprises budget for over the long term. That is our reading of the sector mix, not a strategy the firm has spelled out in full.

Where this fits in the funding landscape

AI venture activity has been dominated by very large rounds for foundation-model developers and heavily capitalized infrastructure players. An $11.3 million fund cannot compete in that arena, and it does not try to.2 Small emerging-manager funds like this one usually compete on access and judgment instead. They try to get into companies early, often before larger firms are paying attention, and lean on the founders' networks to source and support deals.

Stewart's background as a Google vice president and Georges's time at CapitalG give the pair operating and investing experience inside one of the most influential AI companies.2 For a debut fund, that pedigree is likely a key part of the pitch to both limited partners and founders. The fact that the firm built a 10-company portfolio before its final close suggests it found deal flow without the full fund in hand.2

The takeaway

The fund is small, but the thesis captures a widely discussed shift: AI's commercial phase is moving from experimentation to procurement. If that shift arrives on the timeline BAG Ventures expects, early bets on governance, security, and industry-specific software could look well-timed.2

The risk is timing. If enterprise budgets stay in pilot mode longer than expected, startups selling into them may face longer sales cycles and need more capital than seed-stage checks provide. With two more years of deployment planned, the firm will soon learn whether enterprise buyers start paying the way it expects.2