We talk a lot about AI in fertility right now, and most of the conversation is about what AI can do. Write the clinical note, summarise the consultation, answer the patient’s question, draft the letter. All useful. We are building some of it too.
But we think there is a much bigger opportunity.
Your fertility platform needs to learn.
Every fertility clinic already creates an extraordinary amount of data across the patient journey, treatment, medication, embryology, cryo, scheduling, communication, finance and outcomes. The problem is not the amount of data we have, it is that historically we have built systems to store it, not to learn from it.
And there is a pretty big difference between the two.
Today, if a clinic wants to understand why conversion changed, whether a protocol performs differently for a certain patient cohort, where patients drop out, why one embryology workflow produces different results or which operational changes actually improve efficiency, that question often turns into a data project. Someone needs to find the data, connect it, write the query, validate it, put it into a dashboard and then hopefully turn it into a decision.
Fertility care is to some extent digitalised, still not fully. But something we haven’t done is to make it easy to learn from.
The data model matters more than the AI model
I think this gets missed in a lot of the current AI conversation. If your patient data lives in one place, embryology somewhere else, cryo in spreadsheets, financial data in another system and important clinical information is buried in free text, you can put a very clever AI model on top of it, but it still doesn’t really understand how your clinic works.
For AI to become genuinely useful in running a fertility clinic, there needs to be structure underneath it: a patient model connected to a treatment model, connected to the laboratory, finance, scheduling, cryo and outcomes. Not because every clinic should work in exactly the same way, but because the underlying information needs to speak the same language.
This is a big part of what we are building towards at wawa. Not AI sitting on top of the platform as another feature, but intelligence becoming part of the platform itself.
Then you can start asking much more interesting questions
Imagine not needing another dashboard every time you want to understand something new.
You should be able to ask why patients are dropping out between consultation and treatment, which protocols are producing different outcomes for comparable patient populations, where your nurses are spending unnecessary time, why one clinic is converting better than another, which embryology workflows correlate with better blastocyst development, where schedules consistently break down or what actually changed when your pregnancy rates moved.
And importantly, the conversation should not stop at what happened?
It should increasingly move towards why it happens, what can we learn from it and what should we look at next? That is a fundamentally different relationship with your data. The value is no longer just reporting, it is understanding.
This gets really powerful when clinics can learn from each other
A huge amount of knowledge in fertility today is local. A brilliant embryologist learns something in the lab. A doctor changes a protocol and sees better results. An operations team redesigns a workflow and suddenly patients move through treatment faster. One clinic has unusually good conversion, another has figured out how to run its schedule incredibly efficiently.
But most of that learning stays where it happened. I think one of the biggest opportunities in fertility technology is changing that.
When the underlying data is structured in the same way, every cycle becomes another opportunity to learn. You can compare similar patient populations, find variation between protocols, workflows and clinics, understand whether that variation matters and feed what you learn back into how care is delivered.
The loop becomes care → structured data → comparison → learning → change → better care, and then you run it again with the next patient and the next cycle.
That is where scale becomes really interesting to me. More clinics should not just mean more patients and more data. It should mean more learning.
And that is where AI becomes much more interesting
We don’t think the end state is one generic AI chatbot sitting inside your fertility software (or outside).
We believe intelligence will sit alongside the people actually running the clinic. A clinician assistant that understands the patient’s history and treatment, a patient coordinator that knows what needs to happen next, a finance assistant that understands the entire financial journey, an operations assistant that can spot where the clinic is getting stuck, and a business or research analyst that can interrogate years of clinical and operational data through a conversation.
And all of them become much more useful because they are learning from the same underlying platform rather than trying to piece together fragments from the outside.
This is why we keep coming back to something that sounds slightly boring compared with AI: good AI is downstream of good operational data.
We still have a lot of infrastructure to build to get there. Better migrations, better integrations, more structured data, better workflows and much deeper connections between what happens clinically, operationally and financially. But that infrastructure is not separate from the AI strategy. It is the AI strategy.
For years, fertility technology has been built to remember what happened to a patient.
We believe your fertility platform should do something much more valuable with that memory. It should learn.
