Company Visit at Sensi.AI

By: Reuven Dersovitz

I sprint to catch the arrival of the bus to Tel Aviv, as Moovit is showing me that it will be coming much earlier than expected. A short ride later, I arrive in the city with students from the dorms about a two-minute walk from the destination. When we approach what should be the company’s space, we see that there is a nearly inconspicuous hanging sign on an iron gate to an alleyway indicating for us to enter. This alley leads to an unassuming and slightly rickety elevator which we ascend to the floor of the company. But despite feeling skeptical due to the less-than-grand entry of the place, the doors to the company office space open, and we enter a modern millennial’s vision of a workplace: white walls and floors, a large TV surrounded by couches and armchairs, a table topped with cakes and drinks, and office spaces with employees sitting in front of multiple monitors and iPads on stands. We sink into the couches and await the presentation.

The company is Sensi.AI. When the term “AI” comes to mind, at least for me, I think of a new generation utilizing technology in a revolutionary way to self-learn, predict outcomes, and process information in a way that has never been dreamed of before. It’s a young, energizing, and futuristic feeling. However, Sensi takes the use of AI in a different direction. Romi Gubes, the CEO and founder of Sensi.AI, enters the room and sits down with us, and initiates the presentation of her company in a casual manner. She explains the singular product that they offer: an audio recording device that identifies when caretakers, particularly for elderly people either in nursing or private homes, are treating the person they should be caring for in a poor way. This could be by unnecessarily shouting, being rough with objects, or even physical abuse. Sensi does this by first taking a baseline of the audio in the main rooms that the caretaker and the client will be in, and then mapping out irregularities that don’t fit with the baseline. The irregularity then gets sent to an audio annotator who goes over the recording and confirms whether the anomaly is in fact an issue. If it is, the agency of the caretakers or the nursing home is notified in order to take proper action.

But that’s the technical stuff. What was really interesting were the motive, the path they took to get to where they are now, and projections for the future. Romi shares with us an emotional story that displayed the need for Sensi.AI. She once sent her daughter to daycare, and sadly, it turned out that the staff was abusive toward the kids. Romi realized that there was a lack of knowledge of what happens between caretakers and those they are supposed to be taking care of. In fact, she found out that parents sometimes send recorders in their kids' backpacks to daycare and listen to the audio afterward in order to know what happened while their kid was there. So clearly there was an interest in a device of this sort.

Sensi.AI got its start at a nursing home in Israel. However, the product then is certainly not what it is today. It began as a Proof of Concept (POC) in order to confirm that there truly was a need for this product and an ability to make it happen. Romi and her co-founder set up recorders in many of the nursing home rooms, but that is all there was; no code, no baseline, and no hired audio annotators. Instead, the two of them personally listened to every single recording and took note of all the data and potential anomalies in an Excel sheet. This is what we call the “Wizard of Oz'' method, where the backend functions of the product are not set up and are carried out manually while still delivering the same end result. This gives the impression of a fully functioning product without having to go through the entire development process first. The nursing home found the information Romi and her co-founder provided to be very valuable, confirming the value proposition of Sensi.AI. After proving their concept in Israel, they were able to secure funding from prominent venture capitalists and moved on to writing the first lines of code for the development of their Minimum Viable Product (MVP) with the goal of massively penetrating the US caregiving agency market.

When I pushed for the numbers, however, the bigger picture became a little concerning. Sensi has been in business for four years now, and according to Romi, they are projected to break even in a minimum of three years. And that is only if their priority is profitability, which Romi explained isn’t, as she would prefer to prioritize expansion with the hope that in the long run, it will be more beneficial for them. They currently have eighty employees, twenty in Israel and sixty in the US. That is an enormous amount of employees to pay for a company that still has a long way to go, and this is by far their largest cost. At this point, Sensi’s yearly revenue is approximately $4.58 million, but total costs are nearly double that at a staggering $9 million. Romi claims that the company has been growing rapidly at a 3x rate for the past two years and is expected to do it again this year and continue growing. That being said, after this year, their growth is projected to begin to slow down, with the following year at 2x and the years after continuously decreasing. This is indicative of them approaching the peak of the Adoption Curve. Now I’m no financial advisor or projection analyst, but if they are hitting the peak of the curve sometime in the next few years and the break-even point is a minimum of three years away, that doesn’t seem to spell out good news for anyone. There are further concerns that the device only covers specific rooms resulting in gaps in the monitoring when for instance the caretaker takes the individual on a walk outside (perhaps a wearable would solve this problem). In the cutthroat technology world, new innovations or improved-upon ideas quickly render the old ones obsolete. Based on all this, as a potential investor, I would be very hesitant, and as a current investor, I would be anxious.

Even amongst the uncertainty, there is much inspiration to be found. In fact, it is precise because of the uncertainty that admiration exists. The unknown is what startups are all about, and right now Romi is in the thick of it. But her ability to maintain optimism, positivity, loyalty to her dream, and unwavering dedication to the company as a whole when the going gets tough is surely something to look up to. I think the greatest takeaway for anyone interested in going into the startup world is that there is no quick fix, and a high degree of risk tolerance and perseverance is a requirement to reach the light at the end of the tunnel. As a student in the BA Business & Entrepreneurship program, I found it particularly fulfilling to be able to relate some of the concepts we learned in classes to the real world of startups and companies, some of which were mentioned earlier. To conclude, I certainly hope I am way off in my amateur analysis of the company’s future, and that they only have success and continue to provide value to the elderly in need of it.