A Dog Named Rosie Inspired a Cancer Vaccine Startup. The Science Is Still Fuzzy.
Australian entrepreneur Paul Conyngham says AI helped shrink his dog's tumours. Now he has Y Combinator backing and plans for human trials. Researchers say the evidence is far thinner than the pitch.
Key points - Paul Conyngham launched Gamgee in 2025 to offer personalised mRNA cancer vaccines, starting with dogs. - The startup has backing from Y Combinator, one of Silicon Valley's most recognised startup incubators. - Gamgee claims roughly 75% tumour reduction in its first documented case, based on a single dog. - Researchers say it's impossible to credit the mRNA vaccine specifically, because Rosie received a second therapy at the same time. - Gamgee says it is recruiting for clinical trials in Australia, with a global waitlist open now.
A year ago, Paul Conyngham posted on X about using ChatGPT, Grok and other AI tools to design a cancer treatment for his dog Rosie, a Staffordshire bull terrier-Shar Pei mix. Several of her tumours shrank. The internet loved it.
Now Conyngham has turned that story into a company.
What exactly is Gamgee offering?
Gamgee is building personalised mRNA cancer vaccines for dogs.
Conyngham, who openly states on the company website that he is "not a biologist," says AI and genetic analysis helped him tailor Rosie's treatment to her specific cancer. He credits that with adding years to her life. Gamgee wants to do that for other dogs at scale, and eventually for humans too.
The company is working alongside the University of Queensland and research institutes at the University of New South Wales. For vets, it promises the process takes under 20 minutes of their time, delivers a finished vaccine in roughly four weeks, and needs no changes to existing clinic routines. Pricing isn't listed on the site.
Should pet owners be excited or cautious?
Cautious is the right default. Gamgee's headline figure, around 75% tumour reduction, comes from one case: Rosie herself. The company acknowledges outcomes may vary.
The harder problem, reported by The Verge earlier this year, is that Rosie received the mRNA vaccine alongside a separate therapy simultaneously. Two treatments running together leave no clean way to know which one drove the improvement, or whether the combination mattered. Researchers who examined the case said AI received too much credit while the human scientists involved got too little.
That doesn't mean the approach is wrong. Personalised cancer vaccines are a real, active area of research for animals and people alike. It means one dog's improvement is a long way from proof.
What happens next?
Y Combinator's backing carries genuine weight. Clinical trials are recruiting in Australia, a waitlist is open globally, and Conyngham says the company is already "accepting cases globally."
The longer goal is treatments for humans. For a startup whose origin story rests on a single dog whose results can't be cleanly attributed to its core product, that's a big leap. The trials are what the science has to prove itself against, and watching whether Gamgee publishes controlled data will matter far more than its launch video.
If your dog has cancer and you're considering the waitlist, talk to your vet first. Ask what other treatments would run alongside any trial, and how results will be measured.
