Alumni Spotlight: The remote AI-based solution taking care of grandma

Written by

After years as a software engineer working for various companies, Romi Gubes decided to pave her own path. Following a personal incident with her daughter, she realized there was a great void in the supervising and transparency mechanisms of caregiving for the elderly, infants and disabled people.

She then founded Sensi (formerly Clanz) in early 2019, to create a pioneering, audio-based, remote care monitoring solution to detect increased risk of maltreatment of vulnerable, helpless individuals relying on caregivers. Her first entry stop on the entrepreneurship train was at the 5th cohort of 8200 Impact. “It was an incredibly meaningful experience, as all the ideation part of our business was accompanied by the 8200 Impact program and its mentors,” she says. She also met one of her now senior developers in the company through the accelerator, which she considers as an entire ecosystem of advisors, mentors and peers.

Current solutions for monitoring the best caregiving environment include manually going over CCTV footage or household visitations by an official supervisor. As an engineer, Gubes wanted to leverage technology to facilitate a solution that would benefit both patient and caregiver. Using sophisticated AI models based on audio recognition designed for the care environment, Sensi’s expertise lies in the “soft indicators” which can signal possible behaviours. “Our algorithm automatically analyses audio from the caregiving environment -- from shower water to human communications -- and can point at any underlying irregularities,” Gubes explains. The metrics include employee burnout, high-tension sites, and lack of specific training, and results in either positive support to outstanding caregivers, or an immediate alert in case of maltreatment.

And as for privacy? “As an audio-based product it is fundamentally less voyeuristic,” Gubes says. “In 99% of the cases, our indicators focus on the quality of treatment and not on the actual words being said.” This is possible thanks to smart identification of features such as tonality, cadence and intensity, in addition to the consent of all concerned to have the product embedded in the caregiver’s phone or existing home devices.

The COVID-19 outbreak had further accelerated the significance of Sensi’s product, highlighting the blind spots of long-term care providers. “Especially with the elderly, who are isolated in their homes, remote monitoring is now essential,” she says. Investors had noticed, and Sensi has recently secured two million dollars in two funding rounds since the outbreak began. The team of 11 is now working harder than ever on a new pilot with one of the largest home care agencies in the US.