What If Technology Helped Women Know When Their Period Was Not Normal?
For many women experiencing heavy menstrual bleeding, the healthcare journey does not necessarily begin with a doctor. It can begin months or even years earlier with a conversation with a mother, sister or friend, followed by an internet search, an online forum, a period-tracking application and perhaps, increasingly, a conversation with an artificial intelligence assistant.
The GP appointment may come much later, after somebody has already spent considerable time trying to answer a deceptively difficult question about their own body: is what I am experiencing normal?
That question represents an important and potentially overlooked opportunity for women’s health technology. The challenge is not simply giving women more information about menstruation. The challenge is helping somebody understand their own body well enough to recognise when something has changed, assess information from different sources critically and know when seeking professional healthcare is appropriate.
Heavy Menstrual Bleeding Is More Than a Heavy Period
Heavy menstrual bleeding is sometimes reduced to a discussion about the quantity of blood somebody loses, but NICE takes a broader approach. Its guidance emphasises the impact bleeding has on a woman’s physical, social, emotional and material quality of life rather than focusing solely on measuring blood loss.
That distinction matters because somebody does not need to measure exactly how many millilitres of blood they have lost before their experience deserves medical attention. Bleeding that disrupts work, education, sleep, exercise or everyday activities can itself be important information.
Heavy menstrual bleeding can also occur alongside symptoms such as pelvic pain, pressure or bleeding between periods, and there can be multiple underlying explanations. NICE therefore recommends that healthcare professionals take a history that considers the nature of the bleeding, associated symptoms, its effect on quality of life and factors that could influence treatment.
The difficulty is that this clinical history begins long before somebody enters the consulting room.
The Patient Has Usually Been Collecting Data for Months
Imagine somebody whose periods have gradually become heavier.
During the first month, she assumes it is unusual but probably temporary. During the second month, she asks somebody in her family whether the experience is normal. During the third month, she searches online and reads conflicting advice. She downloads an application and begins tracking her cycle. Later, she asks an AI assistant what her symptoms might mean.
Each interaction produces another fragment of information.
Her mother provides family experience. Google provides search results. Online communities provide other women’s stories. A period application provides tracking. An AI assistant provides an explanation generated from the information she enters.
The problem is that none of these sources necessarily creates a coherent clinical story.
By the time somebody eventually speaks to a healthcare professional, she may have months of potentially valuable information scattered across memories, search histories, applications, messages and conversations.
The patient effectively becomes the integration layer between all of those systems.
That is where technology could do something considerably more useful.
We Do Not Need Another AI Doctor
Generative AI creates an obvious temptation for women’s health technology because a chatbot can provide information instantly and at almost no marginal cost.
However, speed is not necessarily the problem that needs solving.
A woman asking an AI assistant whether her bleeding is normal may receive a perfectly plausible explanation. She may also receive generic reassurance, an unnecessarily alarming interpretation or advice that does not adequately reflect her individual circumstances.
The system cannot perform a physical examination, take a blood sample or automatically determine whether symptoms are being caused by fibroids, adenomyosis, endometriosis, a bleeding disorder or another condition.
NICE’s guidance demonstrates why this matters because different histories can lead to different clinical pathways. Depending on symptoms and examination findings, investigations might involve ultrasound or hysteroscopy, while NICE recommends a full blood count for women presenting with heavy menstrual bleeding.
The opportunity for AI should therefore not be to impersonate the clinician.
It should be to help somebody arrive at the clinician better informed.
What If Technology Helped Women Build Body Literacy?
Most health applications are designed around data collection.
They ask when your period started, how long it lasted, how you felt and perhaps how heavy the bleeding was.
That information can be useful, but collecting data is not the same as developing understanding.
A better women’s health platform could help somebody build what might be described as body literacy. Instead of simply recording isolated data points, the technology could help a woman understand her normal patterns and identify meaningful changes over time.
The system might notice that periods have become progressively longer over six months, that fatigue increasingly coincides with heavier bleeding, that pain is becoming more severe or that symptoms are beginning to interfere with work and everyday activities.
The application would not need to diagnose the underlying condition.
Its first job would be helping the user recognise that something has changed.
That difference between tracking and interpretation could become one of the most important opportunities in FemTech.
The Technology Could Build a Body Story
A conventional period tracker might tell somebody that their cycle lasted 29 days.
A body-literacy system could tell a richer story by showing how symptoms have changed over time.
It could record bleeding patterns alongside pain, fatigue, dizziness, sleep disruption, medication, contraception and the effect symptoms have on everyday activities. It could then transform those observations into a timeline that both the user and, with consent, a healthcare professional could understand.
Instead of arriving at an appointment and trying to remember six months of symptoms in ten minutes, somebody could arrive with a structured history showing when changes started, how frequently they occurred and how their quality of life was affected.
That could make technology valuable without asking it to become a doctor.
The application would become a bridge between lived experience and clinical care.
AI Could Help Evaluate the Advice Women Already Receive
There is another problem that technology could address because people rarely receive health information from only one source.
A family member might say that heavy periods run in the family. An online forum might suggest endometriosis. A social-media video might recommend a supplement. An AI chatbot might suggest that the symptoms are normal. A healthcare website might recommend speaking to a GP.
The individual is left trying to decide which source deserves their trust.
A useful AI system could therefore evaluate information rather than simply generate more of it.
Someone could ask whether a claim they encountered online is consistent with established clinical guidance. The system could explain where evidence is strong, where it is uncertain and when the available information suggests that professional assessment would be appropriate.
The objective would not be to tell women what to believe.
It would be to help them understand the quality of the evidence behind what they have been told.
That could become particularly valuable as generative AI makes health information easier to produce but not necessarily easier to verify.
Triage Could Be More Valuable Than Diagnosis
One of the biggest opportunities may therefore lie in triage.
A useful women’s health system does not necessarily need to answer the question of which disease somebody has. It may simply need to help answer what somebody should do next.
Some symptoms might be appropriate to monitor while collecting better information. Others might indicate that somebody should arrange a routine GP appointment. Different symptoms could justify more urgent assessment, while serious warning signs could require immediate medical attention.
Those thresholds should not be invented by an AI model. They should be developed with clinicians, based on established clinical guidance and subjected to appropriate safety and regulatory scrutiny.
The artificial intelligence could help interpret the information, but the clinical pathway underneath it should be transparent and evidence-based.
That distinction changes the product from an AI doctor into a navigation system for healthcare.
Seeking Help Should Feel Like a Rational Decision
There is also a psychological dimension to this problem that technology frequently overlooks.
Someone experiencing unusual bleeding may hesitate to seek help because they are uncertain whether their symptoms are serious enough. They may worry that they are wasting a doctor’s time or that their experience is simply something women are expected to tolerate.
NICE’s public guidance explicitly says that heavy menstrual bleeding can severely disrupt everyday life and that healthcare professionals should take women’s symptoms seriously. It also notes that around one in 20 women aged between 30 and 49 consults their GP each year about heavy periods or menstrual problems.
Technology could therefore do something psychologically important by making the decision to seek care feel evidence-based rather than emotional.
Instead of somebody wondering whether they are overreacting, the application could explain that their recorded symptoms have changed significantly, are affecting everyday life and meet an established threshold at which seeking clinical advice is appropriate.
That is a very different interaction from an AI chatbot simply saying that somebody should contact their doctor if they are worried.
Earlier Care Can Create More Options
The argument that 60 per cent of heavy menstrual bleeding cases ultimately result in surgery should not be used without stronger evidence.
NICE describes a much broader treatment pathway. Depending on the cause, symptoms and individual preferences, treatment can include the levonorgestrel-releasing intrauterine system, tranexamic acid, anti-inflammatory medicines, hormonal contraception and other pharmacological approaches. Surgical interventions such as endometrial ablation or hysterectomy are among the options considered when appropriate, particularly when previous treatments have been unsuccessful or symptoms are severe.
This actually strengthens the case for earlier engagement with healthcare because heavy menstrual bleeding does not automatically lead towards surgery.
There may be several opportunities to investigate, manage and treat symptoms before more invasive interventions become appropriate.
Technology should help women reach that pathway rather than creating a parallel digital pathway that keeps them away from it.
South East London Is Already Changing the Care Pathway
There is an interesting development happening in South East London because the Integrated Care System has established Women’s and Girls’ Health Hubs designed to make specialist support easier to access.
Pilot hubs were established in Bexley and Greenwich and in Lambeth in spring 2025, following feedback from more than 1,400 local women and girls. The services cover conditions including heavy menstrual bleeding, irregular periods, endometriosis, fibroids, adenomyosis, polycystic ovary syndrome and menopause.
Importantly, eligible residents in those areas can self-refer rather than necessarily waiting for a GP to initiate the referral.
That creates an interesting technology opportunity because better digital triage does not need to end with generic advice to contact a doctor.
It could potentially connect somebody to an actual local pathway.
A woman could build her symptom history, receive evidence-based guidance, understand why further assessment may be appropriate and then move directly towards an available healthcare service.
That is considerably more useful than another chatbot conversation.
Trust Could Become the Real FemTech Opportunity
The women’s health technology market has concentrated heavily on tracking.
The next opportunity may be trust.
Users need to trust that the information they receive is clinically grounded. They need to trust that the system will acknowledge uncertainty rather than inventing a confident answer. They need to trust that their intimate health information will not be unnecessarily shared or exploited. They also need to trust that the technology knows when its role should end and the healthcare system’s role should begin.
This creates an opportunity for products built around clinical evidence, privacy, explainability and carefully designed escalation pathways.
A trustworthy system should be capable of telling somebody that it does not have enough information to answer a question safely.
That limitation should not be regarded as a failure of artificial intelligence.
In healthcare, knowing when not to answer can be part of intelligence.
What Could We Actually Build?
The opportunity extends considerably beyond another menstrual-tracking application.
A body-literacy platform could help people understand changes in their own menstrual patterns while a longitudinal symptom record could transform months or years of experiences into a history that healthcare professionals can review. Evidence-checking systems could compare claims encountered through social media, online forums or AI assistants against trusted clinical sources.
Clinically designed triage could help users understand whether symptoms should be monitored, discussed routinely with a healthcare professional or assessed more urgently. Appointment-preparation tools could summarise the questions, symptoms and timeline somebody wants to discuss so that important information is not forgotten during a short consultation.
The same platform could support navigation into local services such as Women’s and Girls’ Health Hubs, while privacy-preserving records could allow users to decide when their information is shared and with whom.
The technology stack already exists to build many of these components.
The more difficult challenge is designing them around trust.
The Patient Should Own the Story
Women’s health information is exceptionally personal.
A menstrual-health application could potentially contain information about bleeding, pain, sexual health, fertility, pregnancy, contraception, mood, relationships and medication.
That creates a responsibility that extends far beyond cybersecurity.
Users should understand what is being collected, why it is being collected, where it is stored and whether it is being shared with anyone else.
A body-literacy platform should therefore be designed around the principle that the individual is building a health history for herself rather than producing a behavioural dataset for somebody else’s business model.
Where information is shared with clinicians or researchers, that sharing should be understandable and appropriately consented.
Trust cannot simply be added to the privacy policy after the product has been built.
It has to be part of the architecture.
The UKBT Institute Opportunity
Heavy menstrual bleeding demonstrates why women’s health innovation cannot be solved by software engineers working alone.
Clinicians understand the medical pathways, while data scientists can identify longitudinal patterns. Behavioural scientists can investigate why people delay seeking care, while UX designers can make complicated health information understandable. Cybersecurity and privacy specialists can protect intimate information, while researchers can evaluate whether the technology actually improves outcomes.
Women themselves bring another form of expertise because they understand what it feels like to live through the uncertainty between noticing that something is changing and deciding that it is serious enough to seek help.
Bringing those perspectives together changes the innovation question.
Instead of asking how artificial intelligence can replace part of the healthcare system, we can ask how technology can make the journey into healthcare earlier, safer and more informed.
That is a considerably more human application of artificial intelligence.
The Bigger Picture
Healthcare technology frequently concentrates on the moment of diagnosis or treatment, but much of somebody’s health journey happens before either of those things occurs.
Symptoms appear gradually. People compare themselves with family members. They search online. They read forums. They use applications. They ask friends. Increasingly, they ask artificial intelligence.
During that period, potentially valuable information is being generated, but it rarely forms a coherent story.
The opportunity is to connect those fragments without pretending that technology can replace clinical judgement.
A well-designed system could help somebody recognise changes in their body, understand the quality of the information they encounter, document what has happened over time and know when professional care is appropriate.
That could make the person arriving at the GP surgery or women’s health hub more informed without asking them to become their own doctor.
The future of FemTech should therefore not be measured simply by how much information we can collect about women’s bodies or how many medical questions an AI chatbot can answer. It should be measured by whether technology helps women understand their bodies, preserve control over their health information and reach appropriate care with greater confidence.
If we can build technology that helps somebody move from wondering whether they are overreacting to understanding why their symptoms deserve attention, then artificial intelligence will not have replaced the doctor. It will have helped the patient arrive at the right door sooner.