Every two minutes, a woman dies somewhere in the world from preventable pregnancy or childbirth complications, according to the statistic UNICEF’s innovation team cites in describing why it backed a new AI maternal health tool. CareMother, developed by Doto Health with support from UNICEF Femtech Ventures, is now deploying in Indian maternity wards, aiming to catch the warning signs of labor complications in real time rather than after a dangerous delay in an overstretched system.
What the software actually automates
CareMother’s core function is to track maternal and fetal vitals continuously and automate two of the most paperwork-heavy, error-prone tasks in labor and delivery: partograph charting, the graphical record clinicians use to track labor progress over time, and cardiotocography interpretation, the reading of fetal heart rate and uterine contraction patterns that can signal distress. Both tasks are traditionally done by hand, under time pressure, by staff managing multiple patients at once — exactly the conditions where small delays or missed patterns can have outsized consequences.
Why automation matters most where staffing is thinnest
In well-resourced hospitals, a nurse might monitor one or two laboring patients at a time. In many under-resourced Indian maternity wards, a single midwife or nurse may be responsible for far more patients simultaneously, making continuous, attentive manual monitoring of every chart and monitor strip difficult to sustain through a long shift. CareMother’s real-time alerts are designed to flag the specific patients who need immediate attention, effectively acting as a triage layer that helps short-staffed teams direct their limited attention where it matters most in a given moment.
A second tool aimed at fragmented records
Beyond labor monitoring, the same development team is building a complementary tool that pulls scattered visits, scans and medication records into a single structured view for clinicians, intended to help during reviews, referrals and shift handovers by highlighting a mother’s medical history, current condition and key events without requiring a clinician to dig through fragmented paper or siloed digital records built up over multiple prenatal visits, sometimes across different facilities.
The funding and backing behind it
UNICEF Femtech Ventures’ involvement gives CareMother institutional backing uncommon for maternal health startups operating in lower-resource settings, where access to capital for women’s health technology has historically lagged far behind funding for other health technology categories. That backing is itself part of the story: femtech investment broadly has been criticized for concentrating on wealthier markets and consumer wellness apps rather than the settings where maternal mortality risk is highest, making CareMother’s India-first deployment something of an outlier in the sector.
The caution researchers want built in from the start
Global health researchers studying AI-assisted obstetric care caution that tools built and validated in one health system don’t automatically generalize to another, and that cardiotocography interpretation in particular has a documented history of high inter-observer variability even among trained clinicians — meaning any AI system trained to automate that interpretation needs rigorous, locally validated testing rather than importing models built on different populations elsewhere. A May 2026 research article on AI in pregnancy care specifically argued that fair adoption depends on explainable models and transparent validation, a standard that is harder to meet in resource-constrained settings where independent auditing capacity is itself limited.
What’s next
CareMother’s rollout will be watched closely as a test case for whether AI-driven maternal health tools can meaningfully reduce preventable deaths in exactly the settings that need the intervention most, rather than remaining concentrated in wealthier health systems that already have more monitoring capacity. UNICEF’s innovation arm has signaled interest in extending similar tools to other high-maternal-mortality regions if early deployment data in India shows measurable reductions in missed complications, making the coming months of real-world use the critical proving ground for the broader femtech-for-global-health model it represents.
Photo: sujins / PIXABAY via Pixabay
