For investors

We placed our own bet first.

Giant Leaf AI is a full-stack foundry for private foundation models — founder-funded to date, with our own capital in the hardware and demand committed before the first rack landed. As the foundry scales, we plan to raise outside capital for the first time. This page is our story.

The story so far

Proof, then hardware, then demand.

01 · Proof

We shipped before we pitched

This team built TalkDoc, an AI behavioral-telehealth company that reached $1.8M peak annual gross revenue with a 0.16% patient hospitalization rate — and published PieBot, research reproducing an AlphaZero-class result at roughly 600× lower training cost.

02 · The bet

$2.3M of our own capital

We bought our 10-system NVIDIA DGX H100 fleet — 80 dedicated GPUs — outright, racking in August 2026. Zero outside capital to date: the founders funded the cluster themselves.

03 · Demand

Committed before delivery

HF0, a premier San Francisco accelerator, committed to 20% of the cluster's capacity in a blended cash-and-equity deal signed before the first rack landed — making us the house model foundry for its startup portfolio.

04 · The flywheel

Every build compounds

Each relationship lands a multi-million-dollar model build and expands into recurring revenue across training compute, our Greenhouse platform, and dedicated hardware. Every model shipped feeds the platform and lowers the cost of the next one.

Where we stand

Capital deployed, capacity committed.

$2.3MFounders' own capital deployed
80H100 GPUs, online Aug 2026
20%Capacity committed pre-delivery
$0Outside capital to date

Rishi Sachdev

Co-founder · Technical & Product

CEO and technical founder of TalkDoc (AI behavioral telehealth). Previously Google Fiber — its first technical program manager, on a top-priority project — and Coinbase; CTO at Rooster. UC Berkeley, CS & Applied Math. Built the MedGemma clinical deployment and PieBot.

David Grossblatt

Co-founder · Business, Growth & Capital

Leads commercial strategy, enterprise partnerships, and pipeline. Drives capital strategy and structures deals that align hardware cost with customer economics — the model behind the HF0 commitment.

The road ahead

Preparing to grow — and to raise.

Every model we ship makes the next one cheaper, and the market is moving our way: industry analysts project enterprise AI to reach roughly $155B by 2030 (Grand View Research), while data-residency rules push regulated enterprises toward private weights.

Planning to raise capital

Giant Leaf AI has been funded entirely by its founders. To expand our owned compute and our research-and-delivery team as customer demand grows, we are planning to raise institutional capital. Any offering of securities, if and when one is made, will be a private placement limited to accredited investors, made only through definitive offering documents under applicable exemptions such as Regulation D — never through this website. If you are an institutional or accredited investor who would like to get to know our team and follow our progress, we'd welcome the conversation.

Introduce yourself

Important notice. This page describes our business and plans and is provided for general informational purposes only. Nothing on this website constitutes an offer to sell, or a solicitation of an offer to buy, any securities of Giant Leaf AI LLC, nor does it constitute investment, legal, or tax advice. No money or other consideration is being solicited through this website, and none will be accepted through it; no offer to buy securities can be accepted, and no part of any purchase price can be received, through this website. Any indication of interest communicated to us is non-binding and involves no obligation or commitment of any kind. Any offer or sale of securities, if one is ever made, would be made only through definitive offering documents, only to persons eligible to invest under applicable law — including, in the United States, "accredited investors" as defined in Rule 501(a) of Regulation D under the Securities Act of 1933 — and in compliance with applicable federal and state securities laws. Statements on this page regarding markets, plans, and expectations are forward-looking, involve risks and uncertainties, and may prove incorrect; historical figures relate to prior ventures or specific engagements and are not indicative of future results. Third-party market estimates are cited for context and have not been independently verified.