What I did wrong with Artemis

Part 2/2 on what didn’t work.

---

In 2024 I started a project that became Artemis, a $35M Focused Research Organization focused on developing better non-mouse animal models for biomedical research. In December 2025, I shut it down to make room for new adventures.

Along the way, I learned a few things that may be helpful if you’re considering a similar journey. So here they are…

---

1. Stubborn on vision and stubborn on the details

FROs are often created on a set of pre-set milestones attached to lots of pre-committed funding. It’s hard to be “stubborn on vision but flexible on the details” when you search for funding in this pre-commitment and technical-first approach. Though it’s now expected of early-stage startups to pivot multiple times, many impact philanthropy modalities, including FROs, are not there yet culturally. When steering a big ship, it’s hard to set the right course, change course, and ultimately arrive at the “destination.” Forest’s “exit” to Merge Labs is incredible. But to this day, it’s still not clear to me what the exit paths are for all FROs. I don’t think there’s necessarily a “correct” path, which is both scary and exciting.

First impressions also matter a lot in philanthropy, and though it took me a while to realise this, I didn’t really want to compromise on the details either. I was searching for a funder who was perfectly aligned. Some call that resilience or courage, others call it naivety. Another way to think about it is that a FRO was not the correct modality for Artemis or me as a scientist/founder.

2. Miscalculated the absolute number of funders

It took too long to realise just how few funders understood and aligned with Artemis’ specific vision. In essence, it was a simple math error where I miswrote the denominator. Though this is the case for arguably all impact philanthropy projects, essentially if you filter by: >$1B liquid + giving today + understands or wants to understand your specific science… you generally end up with less than 10 people in the world.

And then whether you have access to them is a whole other question. Unfortunately, this took me too long to realise.

3. Artemis was poorly framed

VC culture today expects investors to be somewhat technically proficient. Arguably this is a byproduct of efficient markets that forces delta-seeking investors to look where no one else will look. I think that’s a good thing because it makes the founder’s job easier and they’re required to do less prep and homework. But philanthropy usually doesn’t operate this way.

Philanthropists, family offices, and UHNWI already know how to make money. And you’re not selling a new way to get rich either. In impact philanthropy, you must show the funder how they can meaningfully change the world, help their loved ones, and live their life in a more meaningful way through your very, very specific thesis.

I led with the technical details when I should have led with the vision. When I pitched, I was supposed to first help funders imagine the future I saw and the details can be generated as they’re pulled in. If we say VCs do ~50% of their own technical due diligence during a deal, philanthropy would usually do ~10-20%. Philanthropy’s DD is usually done in other ways, which I’ll talk about another time.

4. Artemis was against the zeitgeist

In the age of “more AI,” communicating the nuances of predictive validity and the nuances of animal models were very difficult. My mentors from the ARPAs suggested in late summer that I may not have identified the correct big problem, or, if I had, I wasn’t proposing the right solution.

Both of these things can be true, and it’s something that I’ll continue to reflect on. But if there is quick advice I can pass on, it is:

• Have your thesis align with the current zeitgeist, even if it’s just a “wrapper”.

• Always start from the Forest, then slowly go down to the Chloroplasts. Often, you don’t even need to get to the Chloroplasts to get funders excited.

• Narrative shaping is key, and choose analogies that resonate easily.

5. Solo-founding is hard. Do not recommend.

Co-founders are important not just for the operations but for the emotional rollercoaster of doing something ambitious, often alone. When you’re solo-founding, every single moment you’re not working is another day of gains you forfeit. FROs are a marathon, not a sprint. Find a co-founder, choose carefully, and work together with someone you can really, really, really, really, trust.

6. Artemis wasn’t shut down early enough

I had slowly gained conviction in October that Artemis would not work. This was in large part due to excellent advisors and mentors from philanthropies and the ARPAs that I met at that time. Decreasing your latency is key. Alas, my emotional lizard brain didn’t let me accept that Artemis wouldn’t work until another month or two.

7. Quit the day job

FRO, along with all impact philanthropy, funding timescales are long and uncertain. Impact philanthropy fundraising often feels like you’re trying to grab entropy and bend it the way you desire. Who knows, maybe you’re strong enough. But sometimes, it’s best to wait until the right opportunity arrives. Being a pre-revenue founder is not sexy or fun after some point. A sort of cast and wait approach sometimes actually works.

8. Announcing Artemis publicly did not attract the people we wanted

Here’s the thing: Philanthropists aren’t on LinkedIn.

9. Didn’t build enough in public

Nadia is still a massive inspiration for what I do, and one of the main reasons I publicly wrote these learnings in the first place. One of the biggest regrets, and possibly failures, is that I didn’t write about my process nor the potential of these non-models enough. The sheer grit, resilience, and passion of organismal scientists is simply infectious, and I regret that I still have not told these stories to the world.

10. I have yet to pass the torch.

I started Artemis because I believed no one else would. Not because I was the first to realise mice weren’t great models, but because I saw a new and promising approach called FROs to try creating that change. It didn’t work in this iteration, but that doesn’t mean it was a bad idea. We still need better animal models, especially in complex, multi-system diseases from heart failure to PCOS.

My singular wish is therefore for Artemis to simply exist in the world and the transformative potential of non-model species to be realised. It’s entirely possible that I was not the right person to lead it, but to this day it’s still not clear who can take this project and run with it.

I believe non-models hold the secret to long, healthy lives for humanity. If you believe so too, and want to carry on this project in your own way, do reach out: alexaraki[dot]me[at]gmail.com