Tau Ventures Launches Fund III: Seed Applied AI, Bigger Cheques
TL;DR
Tau Ventures, the Palo Alto seed firm founded in 2019 by Amit Garg and Sanjay Rao, has announced Fund III, its fourth vehicle after Fund I, the Opportunity Fund and Fund II. The strategy is unchanged in shape but bigger in bite: applied AI at seed, across digital health, enterprise and physical AI, with initial cheques stepping up to $500K–$1M from $500K in Fund II and $250K in Fund I. The firm has not disclosed a target size, and the fund's SEC filings were made as indefinite offerings, so a headline number simply isn't in the public record yet. What makes this worth reading is not the headline figure but the shape of the franchise underneath it: 80 companies backed, $93 million-plus in AUM, and 15 IPOs and acquisitions, built by a two-partner team that has quietly run the same thesis through three market regimes.
Key Takeaways
This is a launch, not a close, and the distinction matters. Tau filed Form D for two parallel vehicles, Tau Ventures Fund III, L.P. and Tau Ventures Fund III (QP), L.P., on 14 July 2026. Both list the offering amount as indefinite, with no capital reported sold and first sale recorded as yet to occur as of that filing date. Subscription documents may well have been signed in the days since, and the firm has said LP support is strong. For now, though, the fund is best described as open rather than closed, and a final size has yet to be made public.
The cheque size step-up is the real signal. Going from $250K to $500K to $500K–$1M across three funds is a firm deliberately buying more ownership per position rather than more positions. In a seed market where AI rounds are priced aggressively and party rounds have made ownership meaningless, a $1M cheque is roughly the minimum needed to hold a defensible stake. Tau is choosing concentration over index exposure, which is the correct response to where seed pricing has gone.
Two founding partners and three principals is a genuinely small firm, and Garg says so openly. The announcement ends with an explicit ask for help with sharing and referrals. That is unusual candour from a GP and it tells you the honest thing about sub-$100M funds: the constraint is not capital or judgement, it is distribution and deal-flow surface area. Founders should read that as an invitation, not a weakness.
Parallel QP and non-QP vehicles signal a broadening LP base. Filing a 3(c)(1) fund alongside a 3(c)(7) qualified-purchaser fund lets Tau take both smaller individual LPs and larger institutions without blowing the 100-beneficial-owner cap. That is the structure of a firm expecting institutional cheques it did not have in Fund I, consistent with Garg's note that existing LPs increased and new LPs joined.
Fund Overview
Fund Name: Tau Ventures Fund III (filed as two parallel vehicles: Fund III, L.P. and Fund III (QP), L.P.)
Fund Size: Not disclosed. SEC Form D filings dated 14 July 2026 list the offering as indefinite with no amount reported sold at filing
Stage: Seed, with selective pre-seed; the separate Opportunity Fund handles Series A and B follow-on
Check Size: $500K–$1M initial, with capacity to go larger by syndicating alongside LPs
Geography: United States. Distributed teams welcome, but the CEO needs to sit where Tau can help. Roughly half of Fund I sat in the Bay Area, the rest in Boston, NYC, SoCal and Austin
Focus: Applied AI with a defensible edge (proprietary data, faster training, earlier error detection) across three verticals: digital health, enterprise (HR tech, fintech, cybersecurity) and physical AI (cars, drones, robots). Software and services led, but open to hardware. Prefers companies 9–18 months from their next round
Key LPs: Not disclosed. The firm states existing LPs increased their commitments and several new LPs joined
Firm Stats (per Tau Ventures): $93M+ AUM, 80 companies backed, $6B+ raised by portfolio companies, 15 IPOs and acquisitions
Why This Fund Matters
The interesting thing about Tau is not that it is an AI fund. Every seed fund is an AI fund in 2026, and the label has stopped carrying information. What separates Tau is that it was an AI fund in 2019, before the label was worth anything, and it has kept the same three verticals through the entire cycle: the 2021 mania, the 2022–23 correction, and the current concentration of capital into a handful of foundation-model and infrastructure names.
That consistency has a practical consequence for LPs. A firm on Fund III with a 2019 vintage Fund I is now at the point where Fund I is mature enough to show real DPI rather than paper marks. Tau has not published fund-level performance, and prospective LPs should ask for it directly rather than inferring from the 15 IPOs and acquisitions headline, since that count spans the whole portfolio and does not tell you what came back to Fund I specifically. But the vintage timing is favourable: a 2019 seed fund invested before the peak and had dry powder through the 2022–23 reset, which is close to the ideal deployment window of the last decade.
There is also a structural argument for this size of fund right now. The seed market has bifurcated hard. At one end sit multi-stage giants writing seed cheques as option premiums on Series A access, largely indifferent to seed-stage ownership. At the other sit solo GPs and micro-funds with speed but limited follow-on capacity. A $90–150M firm with a dedicated opportunity vehicle occupies the awkward, and possibly correct, middle: enough capital to lead or co-lead and defend ownership into Series A, small enough that a $200M outcome still moves the fund. Tau's decision to run the Opportunity Fund as a separate vehicle rather than inflating the flagship is the right structural call, because it keeps the seed fund's return math intact while still letting the firm concentrate into winners.
The physical AI vertical is the one to watch. Digital health and enterprise are crowded, well-understood categories with many credible seed investors. Physical AI, meaning cars, drones and robots, is where applied AI meets hardware timelines, capital intensity and real-world safety constraints, and where far fewer seed funds have genuine operating intuition. Tau's portfolio already includes Chef Robotics on the robotics side. If Fund III has an edge that is hard to replicate, it is more likely to sit there than in another HR tech seed deal.
The Team
Tau lists two founding partners on its own team page, and it is worth being precise about titles because third-party databases routinely inflate them. Amit Garg is a founding partner focused on digital health, with some automation and logistics coverage. He spent over 20 years in Silicon Valley including Samsung NEXT Ventures, Norwest Venture Partners and Google, and holds a BS in computer science and MS in biomedical informatics from Stanford plus an MBA from Harvard. Sanjay Rao is a founding partner covering enterprise and automation, previously a program manager at Microsoft, a strategy consultant at McKinsey, and a venture investor at Norwest, and a co-founder of an automation startup. He holds an MBA from Harvard Business School and bachelor's and master's degrees in EECS from MIT, and served five years on the MIT Corporation.
The shared Norwest history is not incidental. Both partners saw institutional venture from inside a multi-stage platform before going out on their own, which is generally a better predictor of durable fund-building than a purely operating background.
Below the founding partners, the firm lists three principals: Sam Bogrov (healthcare; previously M&A at Pediatric Associates, private equity at Investindustrial and Platinum Equity, M&A at Oppenheimer; Wharton MBA), Insoo Chang (enterprise AI and fintech; previously SVP at Citigroup's strategic investment team), and Sharon Huang (previously a software engineer at the Chan Zuckerberg Biohub building protein and genomic sequencing systems; computer science and epidemiology degrees from Stanford). Readers should note these are principals, not partners, on the firm's own site. The SEC filings for Fund III list only Garg and Rao as executive officers of the fund, with Tau Ventures Fund III GP, LLC as the general partner entity.
The firm also maintains an unusually large advisory network spanning healthcare, enterprise and automation, including senior operators and clinicians from Johnson & Johnson, DocuSign, Palo Alto Networks, Workday, Massachusetts General Hospital and Stanford. For a firm of this size, that bench functions as outsourced diligence and customer introduction capacity, and Garg has said explicitly that portfolio companies have won pilots and follow-on investment through advisors and LPs.
Early Portfolio
Fund III is investing into an existing base of roughly 80 companies. Clearly identified seed and early positions on the firm's portfolio page include Chef Robotics, Cerby, ArmorCode, Infinitus, Assort Health, Alaffia Health, Iterative Health, Labelbox, Dasera, CalmWave, Biotia, Causely, Banjo Health, Autonomize, RapidDeploy and Drip Capital, spanning the three verticals. One note of caution for anyone parsing the portfolio page: Tau tags certain entries, including 1Password, Netskope, Oak Street Health and AbSci, as "acquired portco" rather than as direct seed positions, and that label should not be read as a straightforward Tau seed investment in those companies. The firm's own aggregate claims are 80 companies backed, $6 billion-plus raised by those companies, and 15 IPOs and acquisitions.
What This Means for Founders
If you are raising a seed round in digital health, enterprise software or physical AI in the US, this is a fund worth a first meeting, and the qualifying bar is specific enough to self-assess against. Tau wants a distinct applied-AI advantage, not an AI wrapper: proprietary data nobody else can assemble, a training or inference approach that is materially faster or cheaper, or error detection that beats the incumbent workflow. It also wants you 9–18 months from your next round, which is a polite way of saying it wants a credible plan to a Series A rather than an indefinite runway. On health specifically, Garg has been explicit that roughly 95% of the portfolio is unregulated and fast-growing, with limited appetite for FDA oversight and long sales cycles. If you are a Class III device company, this is the wrong door.
The value-add is operator-led and network-led rather than platform-led. There is no 40-person talent team here. What you get instead is two partners with two decades each in the Valley, a large advisory bench that can produce a real clinical or enterprise pilot conversation, and an LP base that has demonstrably written follow-on cheques into portfolio companies. The Opportunity Fund matters more than founders usually realise: it means your seed investor has a structural mechanism to keep supporting you at Series A and B, which is not true of most sub-$100M seed funds and is a genuine reason to prefer a firm like this over a larger fund that will treat your seed as a cheap option. The trade-off is that a $500K–$1M cheque will rarely lead your round on its own, so plan the syndicate accordingly.
Fund Momentum Take
We like this franchise more than the absent headline number suggests, and the reason is boring: consistency. Tau has run the same thesis, in the same three verticals, at the same stage, since 2019, and has now raised through it four times. In a seed market where a great many firms rewrote their decks around AI in 2023 and will rewrite them again around whatever comes next, a firm that did not have to change anything has a credibility that does not show up in a fund-size headline.
The honest risks are two. The first is that the size that makes the strategy elegant also caps the outcome. On roughly $93 million of AUM with $500K–$1M initial cheques, Tau needs genuine outlier exits to return a fund several times over, and it is buying into 2026 seed valuations to get them. Ownership discipline helps, but the entry-price environment is unforgiving and the firm cannot price its way out of a bad round the way a $3 billion fund can. The second is key-person concentration. Two founding partners doing the investing, however good, is a real diligence question for institutional LPs, and the answer will determine whether Fund III can attract the kind of endowment and foundation capital that turns a good small firm into an enduring one. The parallel QP vehicle suggests they are pursuing exactly that.
Our bet: Fund III lands in the $75–150M range and does fine, with returns driven by two or three positions rather than portfolio-wide performance, which is how seed is supposed to work. The broader read for GPs currently in market is the one worth internalising. Tau raised a fourth vehicle without a splashy number, without a mega-brand anchor LP, and largely on the strength of existing LPs increasing their commitments. In this fundraising environment, that re-up rate is the metric that actually matters, and it is the one prospective LPs should be asking every emerging manager to disclose.
Frequently Asked Questions
How big is Tau Ventures Fund III?
Not publicly disclosed. Tau filed Form D for two parallel Fund III vehicles on 14 July 2026, both listing the offering amount as indefinite with no capital reported sold and first sale yet to occur as of the filing date. Any specific figure circulating should be treated as unverified until the firm confirms it or an amended filing appears.
Has Fund III closed?
No. The firm has announced the fund and indicated that existing LPs increased their commitments and new LPs joined, but this is a launch announcement rather than a final close, and the SEC filings are consistent with a fund at the start of its raise.
What does Tau Ventures invest in?
Applied AI at seed stage across three verticals: digital health (predominantly unregulated, fast-growing companies), enterprise software (particularly HR tech, fintech and cybersecurity), and physical AI (cars, drones, robots). Software and services led, but the firm describes itself as hardware-friendly.
What is the cheque size, and how has it changed?
Fund III targets $500K–$1M initial cheques, up from $500K in Fund II and $250K in Fund I. Tau can go larger by partnering with its LPs, and runs a separate Opportunity Fund to double down on the seed portfolio and selectively invest at Series A and B.
Who runs the firm?
Amit Garg and Sanjay Rao are the founding partners and the only two executive officers named on the Fund III SEC filings. Sam Bogrov, Insoo Chang and Sharon Huang are listed as principals on the firm's team page. The firm is based at 555 Bryant Street in Palo Alto, with presence in DC, NYC, LA and SF.
Have a fund closing to announce? Submit your fund here.
Need help raising capital? Check out our Fundraising Advisory services.