X

Nanni AI Knows Why Your Baby Is Crying. Here's How It Does That

The AI-powered translation app has learned to distinguish cries.

Lisa Lacy Lead AI Writer
Lisa joined CNET after more than 20 years as a reporter and editor. Career highlights include a 2020 story about problematic brand mascots, which preceded historic name changes, and going viral in 2021 after daring to ask, "Why are cans of cranberry sauce labeled upside-down?" She has interviewed celebrities like Serena Williams, Brian Cox and Tracee Ellis Ross. Anna Kendrick said her name sounds like a character from Beverly Hills, 90210. Rick Astley asked if she knew what Rickrolling was. She lives outside Atlanta with her son, two golden retrievers and two cats.
Expertise Technology | AI | Advertising | Retail
Lisa Lacy
4 min read
Screenshots from the Nanni AI app highlight some of its functionality.
Ubenwa Health/Screenshot by Lisa Lacy/CNET

It's late at night. You haven't had a good night's sleep in weeks. And your baby is crying again. Historically, it's been up to you to cycle through the diaper, bottle and/or comfort-object routine with fingers crossed for a fast and easy solution.

AI Atlas art badge tag

Now, however, a startup called Ubenwa Health has an AI-powered translation app for babies, which seeks to help caregivers get to the root of their child's discomfort even faster. It's called Nanni AI and it's available for free for both iOS and Android devices.

Nanni AI is kind of like Shazam for baby cries. When an infant is upset, you can play the child's cry for a few seconds and the app will tell you the most likely reason, as well as provide recommendations for soothing.

Cries are divided into four categories: hunger, discomfort, emotion and pain. Tips for soothing are ordered based on what the app hears in the cry, so caregivers know which to try first and can quickly move to the next recommendation if they've already tried something. This is all thanks to training data from real-world babies.

Training data is what makes tools like popular chatbot ChatGPT work. It's what AI models ingest to make sense of future queries. For chatbots, it's typically written material like books and online posts so they can better understand language. But for speciality AI tools like Nanni AI, the training data is different.

Co-founder and CEO Charles Onu began research in baby cries in Nigeria before moving to Montreal to get a master's degree and then a doctorate in machine learning from McGill University. There, he met co-founder Samantha Latremouille and Ubenwa was born. Their original focus was to use crying to detect the risk of neurological disorders in newborns. ("Ubenwa" means "the cry of a baby" in the Nigerian dialect Igbo.)

Ubenwa began clinical studies at hospitals in Canada, Nigeria and Brazil about four years ago. As part of research, it began collecting cry recordings, along with why the baby was upset if there was a clear reason, such as the baby just got a shot or was due to be fed and stopped crying afterward.

"We've collected thousands and thousands of cries through the clinical studies and then trained algorithms to try and detect and understand which cry is which," said Florent Voumard, business operations lead at Ubenwa.

The startup released its cry detection app in February. With parents using it, Ubenwa's baby cry database has grown to 1.5 million cries and is expanding by tens of thousands of cries daily. The addition of parent feedback, such as the baby was too hot or too cold or needed a diaper change, helps Ubenwa refine its algorithms. (The app has about 150,000 downloads to date.)

The goal is to help improve babies' — and parents' — sleep and overall well-being.

Nanni AI also offers tracking tools for sleep, feeding and diaper changes.

"We want to be a bit of a bridge between parents and the pediatrician for those first visits," Voumard said.

The hope is the app makes it easy for even frazzled new parents to record baby's sleep, feeding and diaper routines by either talking to the app, which records events for future summaries, or by entering the information manually.

Eventually, Ubenwa wants to help parents predict the best time to put babies to bed, as well as when they're likely to wake up, based on their routines. But the startup doesn't have a release date for this functionality.

Further down the road, Ubenwa aspires to turn baby cries into the next vital sign as part of a remote health monitoring system for newborns, and perhaps even to be able to detect medical abnormalities in infant cry sounds. But first it needs approval from Health Canada and the US Food and Drug Administration.

In July 2022, Ubenwa announced a $2.5 million preseed financing round led by AI-focused investment firm Radical Ventures. AIX Ventures, AI pioneer Yoshua Bengio, and Google Brain's Hugo Larochelle and Marc Bellemare also participated.

"AI is well-suited to deriving insights from the sound signature of an infant cry," Bengio said in a statement at the time. "Onu's leading research into identifying biomarkers in baby cry sounds offers the promise of unlocking our understanding of what's behind a baby's cry."

This is one of a series of short profiles of AI startups, to help you get a handle on the landscape of artificial intelligence activity going on. For more on AI, see our new AI Atlas hub, which includes product reviews, news, tips and explainers.

Editors' note: CNET used an AI engine to help create several dozen stories, which are labeled accordingly. The note you're reading is attached to articles that deal substantively with the topic of AI but are created entirely by our expert editors and writers. For more, see our AI policy.