Venture Capital

Updated:

Gradient Ventures

Gradient Ventures, founded by Anna Patterson inside Google, runs an AI-only early-stage fund backed by Alphabet's balance sheet.

Gradient Ventures logo

Gradient Ventures

Formed in 2017, Gradient was the brainchild of Anna Patterson, a former VP of engineering at Google whose deep machine-learning background shaped the fund's DNA. Rather than spin out as a standalone partnership, Gradient kept a unique tether: it operates as an independent entity yet invests solely off Alphabet's own capital. The founding premise was straightforward — apply Google's internal AI expertise to external startup investing, giving the firm a technical diligence edge that pure-finance VCs find hard to replicate. Gradient's strategy centers on seed to early-stage checks, typically sizing $1M to $8M, across applied AI verticals including enterprise software, fintech, cybersecurity, and robotics. The firm avoids pure-play research labs, favoring startups with a clear product surface and recurring revenue. Portfolio companies confirmed include KitCheck (hospital pharmacy automation), Arena (customer analytics), and Algorithmia (ML infrastructure) prior to its sale to DataRobot. Gradient participates in both solo-led rounds and syndicates alongside traditional firms like Sequoia and a16z. Geographically, the team covers North America chiefly from San Francisco, though it evaluates companies globally when Google's cloud or product signal surfaces them. Gradient's team blends operators and investors. Anna Patterson runs the firm with co-founding General Partner Zachary Bratun-Glennon, a longtime Google alumni and deal lead. The wider partnership includes engineering PhDs and ex-product leads, which enables Gradient to staff a technical fellows program rather than a large post-investment services platform. In 2024, the firm participated in the $25M Series A of CentML, a Toronto-based AI infrastructure startup, signaling continued appetite for machine-learning tooling. Though exact AUM remains undisclosed, secondary estimates from the venture community place invested capital north of $200M across multiple vintages. What genuinely sets Gradient apart is its structural symbiosis with Alphabet — no other corporate venture fund operates with this degree of independence yet exclusive LP commitment. Gradient is not Google's strategic arm; it invests purely for financial return. That posture means portfolio companies get the technical advisory of Google's engineer bench without the expectation of eventual acquisition. That arm's-length relationship, combined with an AI-only mandate, makes Gradient more thesis-concentrated than peers like GV (also Alphabet-backed but sector-agnostic) and more purely financial than Intel Capital.

General information

Firm type

Venture Capital

Year founded

2017

AUM

Undisclosed

Location

Region

North America

Country

United States

City

San Francisco

Corporate office

San Francisco, CA, United States

Principals

Anna Patterson

Founder and Managing Partner

Zachary Bratun-Glennon

General Partner

Sector focus

AI/MLEnterprise SoftwareFinTechCybersecurity

Frequently asked questions

Who runs investment decisions at Gradient Ventures?

Investment decisions are led by Founder and Managing Partner Anna Patterson alongside General Partner Zachary Bratun-Glennon. Patterson, a former VP of Engineering at Google, built the firm's AI-centric thesis and maintains final say on allocations. The partnership operates with a flat structure, relying on technical fellows rather than a large junior team for diligence.

How is Gradient Ventures different from GV or CapitalG?

While all three invest Alphabet's capital, Gradient is the only one with a mandate exclusively focused on applied artificial intelligence. GV is sector-agnostic and stage-flexible, and CapitalG concentrates on late-stage growth. Gradient also functions as an independent entity rather than a corporate division, giving it operational autonomy from Google's product strategy.

Does Gradient Ventures require portfolio companies to use Google Cloud?

No. Gradient is a purely financial-return investor, not a strategic corporate development arm. Portfolio companies receive optional access to Google's engineering expertise and cloud resources, but there is no contractual obligation to adopt Google technology or structure toward a future acquisition by Alphabet.

What investment stages does Gradient typically target?

The firm writes seed and Series A checks, typically ranging from $1 million to $8 million, into startups with a shipped product and initial customer traction. Gradient occasionally follows on into later rounds but does not lead growth-stage deals. The geographic focus is principally North America, with selective earlier-stage participation in other regions when compelling AI signals emerge.

How does Gradient source deals?

Sourcing blends a traditional partner network with a unique internal signal pipeline. Because Gradient operates with Alphabet's data infrastructure proximity, the team can surface startups demonstrating technical heft well before conventional VC databases catch them. The partnership also leans on Google's deep engineering bench for technical reference checks during diligence.

Is Gradient's capital structure a traditional venture fund?

No. Unlike most VC firms that raise blind-pool funds from external limited partners, Gradient invests off Alphabet's own balance sheet under a committed capital model. This single-LP structure eliminates fundraising cycles and allows for faster decision velocity, though Alphabet does not disclose specific vintage sizes.

Which sectors does Gradient explicitly avoid?

Gradient does not invest in biotechnology, pure hardware plays, or defense-focused AI applications. The firm sticks to applied AI within enterprise software, automation, developer tools, and select financial-technology verticals. It also avoids companies whose core business is content generation or social media without a deep technical moat.

Profile maintained by using OSINT (open-source intelligence), regulatory filings, licensed data partners, and verified direct submissions. Read the methodology. Last updated: . Continuous refresh with full update cycles at least every 30 days.

Need institutional-grade insight on venture capital firms?

Altss delivers:

Principals with verified direct contactsAllocation history by asset classOSINT-derived deal signals
Book a demo

Prefer a guided tour?

We’ll walk you through:

Interactive funding timelinesCustom mandate & allocation filters
Book a demo

Browse by category

More San Francisco Venture Capital profiles