It is 6:47 a.m. inside an IVF lab in Northern California. An embryologist named Dana leans into a microscope, eyes adjusting to a five-day-old blastocyst the size of a poppy seed. On the screen beside her, a piece of software ranks the same embryo against thousands of others it has already seen. The two opinions agree. Dana exhales. The software is called Alife.
For four decades, this scene played out without the second voice on the screen. Embryos were graded by eye, by experience, by the unspoken art of pattern recognition built up over careers spent peering into petri dishes. The grading was excellent. It was also, by definition, inconsistent. Different labs, different days, different humans, different answers. Alife Health, founded in 2020 and headquartered on Buchanan Street in San Francisco, exists for one stubborn reason: to give that art a memory, a baseline, and a benchmark.
01The problem they saw
In-vitro fertilization is among the most studied procedures in modern medicine and, paradoxically, one of the least standardized. Pregnancy rates per transfer vary widely by clinic, by patient age, by protocol. The data that could explain why is generated everywhere - and stored, in many places, on paper. Patients drop tens of thousands of dollars per cycle. Insurance coverage is patchwork. The emotional cost is in a category of its own.
Paxton Maeder-York, the company's founder, had been working in surgical robotics at Auris Health before its $5.75 billion acquisition by Johnson & Johnson. He noticed something obvious to engineers and inconvenient to almost everyone else: fertility labs were running 21st century biology on 1990s workflows. Spreadsheets. Faxes. The occasional shared drive named, optimistically, "FINAL_v3."
02The founder's bet
Maeder-York's wager was not subtle. Take the messiest, highest-stakes, most emotionally charged corner of healthcare and apply machine learning carefully - publishing the research, earning the regulators' trust, partnering with the clinicians who would actually use the thing. No "disrupt the embryologist" pitch deck. No replacing the human in the white coat. The bet was that fertility specialists, given a well-built second opinion, would do better work and offer patients better odds.
Investors agreed. Lux Capital led a $9.5 million seed round in 2021, with backing from a who's who that included Anne Wojcicki of 23andMe and Fred Moll, the founder of Intuitive Surgical and Auris. In March 2022, Alife closed a $22 million Series A co-led by Lux, Union Square Ventures, and Maveron. Total raised to date: $31.5 million. The cap table reads like a small museum exhibit on medical device history.
03The product, in plain English
Alife sells software to IVF clinics. The platform is called Alife Assist, and underneath the brand live a handful of modules that each take a piece of the fertility journey and sharpen it. Embryo Assist captures and grades embryo images and turns the day-five report - usually a hand-written sticky note in some labs - into a clean EMR-connected record. Stim Assist suggests medication dose and timing during ovarian stimulation, the phase that determines how many eggs make it to retrieval. Schedule Predict tries to fix the perpetual lab traffic jam. Embryo Predict ranks blastocysts by predicted viability and now carries a CE Mark under the European Medical Device Regulation, a paperwork mountain no one climbs for fun.
There is also a patient app, available on the App Store, that does the modest but useful work of tracking egg retrieval and embryo transfer milestones. Modest, because the company has resisted the urge to gamify the most painful weeks of a patient's life with streaks and badges. Useful, because the rest of the experience tends to be a PDF emailed by a nurse on a Tuesday.
A short company chronology
Five years. Two continents. Several thousand embryos quietly re-evaluated by a piece of software with a soft logotype.
04The proof
Most healthtech startups talk about pilots. Alife has them, and has graduated from a few. US Fertility - one of the largest fertility networks in the country - chose Alife's embryo image capture and reporting platform for nationwide deployment in early 2025. Innovation Fertility signed on to bring the AI toolkit to clinics across the United States. The company publishes peer-reviewed research alongside its product releases, which is rare for a Series A startup and tells you something about who is running the place.
Where Alife shows up in the IVF workflow
Approximate module footprint across the patient journey
Bars indicate relative module depth within the IVF cycle, not market share. Designed to be read in ten seconds.
05The mission, said simply
Alife's stated goal is to modernize and personalize IVF with AI. The unstated goal, audible in every press release, is to lower the cost and increase the success rate of a procedure most people only consider when they have already exhausted easier options. Fertility care is in something its own backers have called a crisis. One in six couples globally faces some form of infertility. IVF success rates have climbed for decades, but plateaued in patches. Most clinics in the United States still operate as independent shops with their own protocols, their own data silos, and their own learning curves.
06Why it matters tomorrow
The interesting bet inside Alife is not that AI will outperform humans in the lab. It is that consistency will. A model trained on tens of thousands of embryos remembers what a tired Tuesday morning forgets. It also creates a shared vocabulary that lets one clinic learn from another. Multiply that across networks, across continents, across a CE Mark and a few hundred more peer-reviewed papers, and you get something the field has never had: a baseline. Not a verdict. A baseline.
There are risks. Regulators move slowly. Embryologists are, justifiably, particular about who gets to grade their embryos. Patients have heard "AI" promises before, and they are skeptical, and good for them. Alife's answer so far has been to publish the work, ship the product, and let the data speak. That is a slower path than the one a louder competitor might take, and it is probably the right one.
Back to the Northern California lab. It is now 7:14 a.m. Dana has finished the morning's grading - thirty-one embryos, five patients, one cup of cold coffee. The software's rankings sit next to hers in a shared workbook the clinic now treats as a quiet medical record. Patients will get cleaner reports later today. The lab will run a hair faster next week. None of it is theatrical. All of it is, by the numbers, the point.