Vivodyne's AI-Enabled Lab Is Growing Vascularized Tissues for Pharma Clients

The Penn spinout's 23,000-square-foot facility aims to replace animal testing with AI-scale experiments on lab-grown human tissues.

About Vivodyne

Published

Vivodyne's robotic arms are dosing lab-grown human liver tissue with a chemotherapy drug right now. They have been for hours. In a 23,000-square-foot facility in South San Francisco, the startup is running what it calls a "human data center," a high-throughput lab designed to generate predictive biological data at a scale that, the company claims, already outpaces all animal trials in the United States [TechCrunch, August 2026]. It is a $82 million bet that the future of drug discovery is not in silicon alone, but in silicon controlling biology.

Founded in 2021 as a University of Pennsylvania spinout, Vivodyne sells its platform to pharmaceutical and biotechnology companies. The wedge is straightforward: replace or supplement unreliable animal models and slow conventional lab work with automated experiments on functional human tissues. The system combines microfluidics, robotics, and AI to grow tissues, dose them, and analyze the results [Technical.ly, April 2021]. The goal is to give drugmakers more accurate human data before a candidate ever enters a clinical trial, potentially saving years and billions in development costs [Vivodyne, November 2023].

The bet on a human data center

The core of Vivodyne's proposition is throughput married to biological relevance. The company's HIVE system is reported to grow 20 different types of human tissue, from liver and airway to bone marrow. Robots autonomously dose these tissues with drug candidates and record the responses across hundreds of thousands of parallel experiments [Superpower Daily, retrieved 2026]. This generates the causal, time-series data that AI models need but often lack, moving beyond correlations found in static datasets [Superintelligence News, retrieved 2026].

The company opened its flagship facility in August 2026, calling it the world's largest human data center [TechCrunch, August 2026]. The scale is the product differentiator. Vivodyne states its platform achieves "twice the throughput of all animal trials being held in the US" [AIMidday, retrieved 2026]. For buyers, the appeal is velocity and predictive power. Early, non-peer-reviewed accuracy claims cited by the company are ambitious: 94% predictive accuracy for liver toxicity, 96% for airway tissue behavior, and 100% concordance when testing 20 chemotherapy drugs on bone marrow [Techticia.com, retrieved 2026].

The academic engine and venture backing

The technology was born in an academic lab, a common pedigree for deep tech bets. Co-founder and Chief Science Officer Dan Huh is a professor at the University of Pennsylvania and directs the BIOLines Lab, where the underlying organ-on-chip and tissue engineering work was developed [Technical.ly, April 2021]. Co-founder and CEO Andrei Georgescu earned his Ph.D. in Huh's lab, commercializing the research [Vivodyne, November 2023]. This academic foundation has been a key signal for investors navigating the high-risk, high-capital-intensity world of lab hardware.

Vivodyne has raised $82 million across three primary rounds. Khosla Ventures, known for backing ambitious science-driven companies, has led the last two, including a $40 million Series A in May 2025 [STAT, May 2025]. The investor syndicate is a mix of deep tech and life science specialists.

Date Round Amount Lead Investor
April 2021 Seed $4 million Kairos Ventures [Technical.ly, April 2021]
November 2023 Seed $38 million Khosla Ventures [Vivodyne, November 2023]
May 2025 Series A $40 million Khosla Ventures [STAT, May 2025]

The 2021 round was meant to grow the team from six to about 18 people [Technical.ly, April 2021]. By March 2026, headcount had reached 43 employees [Revelio Labs, March 2026], indicating a steady build-out of technical and commercial operations.

The regulatory and commercial runway

Vivodyne is not operating in a vacuum. A significant tailwind comes from regulators. The FDA and NIH are actively pushing for alternatives to animal testing, a movement that creates a receptive environment for human-relevant data [Endpoints News, retrieved 2026]. The company's focus on creating "preclinical certainty" aligns with an industry trend where drugs are becoming more precise and human-specific, making animal models less applicable [Dealroom.co, retrieved 2026].

Commercially, Vivodyne is targeting two main partnership avenues: new therapeutic program development and providing supporting data for drugs already in Phase I or II trials, potentially rescuing underperforming assets [Longevity.Technology, retrieved 2026]. The company has discussed working with "several" pharmaceutical partners but has not publicly named them [Forbes, November 2023]. This is typical for early-stage biotech tools providers where deals are often confidential.

Where the model faces pressure

The ambition is vast, and so are the execution risks. Building and scaling a robotic wet lab is a capital-intensive endeavor with complex biological and engineering hurdles. The company's published accuracy figures, while impressive, are from internal studies and require broader validation across more drug classes and tissue types. Furthermore, the sales motion involves displacing entrenched, though flawed, workflows in a conservative industry.

  • Biological complexity. Growing functional, vascularized tissues that accurately mimic human organ responses over long durations is a frontier science challenge. Scaling it robotically adds another layer of engineering risk.
  • Commercial adoption. Pharmaceutical companies are slow to change core R&D processes. Convincing them to trust a new data source for million-dollar decisions requires not just technology, but extensive validation and a proven track record.
  • Capital intensity. The $82 million war chest is substantial, but building and operating advanced robotic labs is expensive. The path to profitability likely requires signing large, multi-year enterprise contracts to cover the high fixed costs.

The company's answer to these risks is its scale and data generation thesis. By producing more human data, faster, it aims to create a feedback loop that improves its own models and becomes indispensable to partners. The recent facility opening is a tangible step toward proving that industrial-scale operation is possible.

The next validation milestone

For a company at this stage, the next twelve months are about translating its technical showcase into commercial proof points. The key milestone to watch is the announcement of a named, flagship pharmaceutical partnership. A public deal with a top-20 pharma company would serve as a powerful endorsement, providing revenue validation and a case study for the broader market.

Internally, scaling the team from 43 employees while managing burn against the $40 million Series A will be a focus. Externally, presenting more peer-reviewed or partner-validated data on predictive accuracy will be crucial for building scientific credibility beyond venture circles.

Khosla Ventures has now backed the company twice, leading a $38 million seed in 2023 and the $40 million Series A in 2025. They are joined by Bison Ventures, MBX Ventures, and CS Ventures, among others [Vivodyne, November 2023] [STAT, May 2025]. The question for that syndicate, and for the biopharma industry watching, is whether robotic throughput can finally crack the code on predictive human data at the outset of the discovery process.

Sources

  1. [Technical.ly, April 2021] Vivodyne is a University of Pennsylvania spinout developing robotic, AI-enabled laboratories | https://technical.ly/startups/vivodyne-testing-raise/
  2. [Vivodyne, November 2023] Vivodyne announces $38 million seed financing led by Khosla Ventures | https://www.vivodyne.com/blog/vivodyne-announces-38-million-seed-financing-led-by-khosla-ventures
  3. [STAT, May 2025] Vivodyne raises $40 million for organoid research | https://www.statnews.com/2025/05/29/animal-testing-alternatives-biotech-vivodyne-raises-40-million-for-organoid-research/
  4. [TechCrunch, August 2026] AI isn’t close to curing cancer, this startup says it knows what it will take | https://techcrunch.com/2026/08/19/ai-isnt-close-to-curing-cancer-this-startup-says-it-knows-what-it-will-take/
  5. [Forbes, November 2023] This Startup Raised $38 Million To Fill One Of The Biggest Gaps In Drug Discovery | https://www.forbes.com/sites/alexknapp/2023/11/22/this-startup-raised-38-million-to-fill-one-of-the-biggest-gaps-in-drug-discovery/
  6. [Superpower Daily, retrieved 2026] Vivodyne's HIVE system grows 20 types of tissue | Not available
  7. [Superintelligence News, retrieved 2026] Vivodyne's robotic HIVE labs produce causal data | Not available
  8. [AIMidday, retrieved 2026] Vivodyne's human data center is achieving twice the throughput of US animal trials | Not available
  9. [Techticia.com, retrieved 2026] Vivodyne's liver cells show 94% predictive accuracy | Not available
  10. [Endpoints News, retrieved 2026] Vivodyne grows vascularized human tissues as FDA pushes alternatives | Not available
  11. [Dealroom.co, retrieved 2026] Vivodyne focuses on creating preclinical certainty | Not available
  12. [Longevity.Technology, retrieved 2026] Vivodyne's platform is commercially available for partnership | Not available
  13. [Revelio Labs, March 2026] Vivodyne had 43 employees as of March 2026 | Not available

Read on Startuply.vc