The company was one of three winners of the Innovate.AI competition, taking the grand prize for the Israel region ahead of about 200 other startups. ZenCity is based in Tel Aviv, but has about 20 city customers in the U.S. and Israel.
The $1 million funding is, relatively speaking, a pretty big deal for ZenCity. The company, launched in 2015, had only raised $1.7 million before winning the competition, so the prize money represents a big chunk of its total funding.
“The million-dollar prize emphasize how serious this issue was for Microsoft,” said Eyal Feder, chief executive officer and co-founder of ZenCity. “This wasn’t a $100,000 competition where everyone applies. It shows real commitment.”
The contest, according to a blog post on Microsoft’s website, is meant to drive adoption and improvement of artificial intelligence.
And AI is a central component to what ZenCity does. The company pulls together multiple data streams containing clues about what problems citizens have, what they care about and how they feel about things a local government can control — from emails, customer relationship management platforms, telephone hotlines, social media and other places. Then it runs them through four algorithms to help the city understand them.
- Classification: Determining what the feedback is about and who in city government a piece of feedback is relevant to.
- Sentiment: Figuring out whether the feedback is positive or negative.
- Anomalies: Identifying trends, like when people are suddenly talking more about a certain subject, or if sentiment has rapidly shifted.
- Geolocation: Pinning feedback to a location so that the city can see in which neighborhoods different issues or discussions are most relevant.
“Generally speaking, if we agree with the pretense that managing a city is very hard, then I think we can agree that one of the [main problems] is prioritizing,” said Feder.
It also has the potential to alert cities to problems they otherwise wouldn’t have known about. That’s because the product features an alerts system based on anomaly detection.
“If there’s a sudden increase in negative sentiment around street cleaning, maybe something is going on there,” he said.
For example, Feder said one city in Israel used the tool to figure out where to spend its budget for fixing cracks in sidewalks. But as it searched for sentiment on sidewalks, it found that there were a lot more people talking about sidewalks being blocked than there were people talking about cracks.
“They didn’t know that was such a major problem in the city, and what they did was shift a lot of that budget from cracked sidewalks to [unblocking] sidewalks,” he said.
And then they watched as positive sentiment built around the work the city was doing to clear sidewalks for public use.
Another possible use for the technology is combatting misinformation. Feder said another city using the company’s product once found a sudden spike in people using the word “murder” in a certain area. But what they were talking about wasn’t a murder — it was an accident. So the city moved quickly to get the word out about what happened, and people stopped spreading the misinformation.
For now, city managers are some of the most common users of ZenCity. But in the future, Feder said, the company might expand to other levels of government.
“Right now all our clients are local governments,” he said. “That’s not to say that we won’t have states or federal governments in the future, but right now we’re focused on the unique [aspects of] cities.”
