Spotlight on the CyRise teams: badook
A pivot: the reaction to the realisation that no one wants to buy the product you’ve built. badook knows this all too well. And we put it bluntly and honestly, because they themselves are blunt and honest. Yaniv Rodenski and Yariv Triffon are our Israeli duo, with a tenacity, sense of humour, and drive that is unparalleled. Their product, which protects AI and Machine Learning models from bad data, is the child of a pivot. Given their vision and their traction, it’s clear that the pivot was the right move.
Yaniv gives us the backstory. “For five months we’d been working on a different idea. We did a lot of work and a lot of pitching, that’s when we first met with CyRise.” As part of the pitching, Yaniv was interviewed for CyRise Cohort 2 in 2018, pitching his original idea, which wasn’t a fit for the program at the time. “Every time we’d talk about this thing, it garnered more interest. What we noticed, though, was that in our three-pronged approach, people focussed on just one part: the testing of the data aspect of the solution. And that became badook.”
It’s fair to say that Yariv was not sold on the idea. “Well, I remember things differently,” he laughs. “We were on a call, waiting to chat to the CTO of a large VC when Yaniv tells me we’re going to dump the old idea and pitch the new one. My response? ‘Dude, we’re going to dump everything we’ve worked on until now?!’ But the moment we brought it up on that call, that was all they wanted to talk about.”
Yaniv had adopted that startup mentality. “I’d been proudly having more conversations with past founders. I’d heard a lot from their experience about pivoting. A pivot was a natural thing.”
After the impromptu new pitch to the VC, and with five months of code down the drain, it was time for Yaniv to return from Sydney to Melbourne. Yariv, who is based in Sydney, was dropping him to the airport. In the car they debated the pivot decision. Now, we don’t know for sure, but it’s fair to assume this was a tense car trip. It ended with Yaniv telling Yariv to listen to the Master of Scales podcast, scrambling out of the car curbside, and jumping on his flight home. On the drive home after dropping off Yaniv, Yariv listened to the podcast. That was enough to convince him. He sent a message to Yaniv while he was mid-flight. It read: “Let’s do it.”
“I wasn’t aware of how normal it was,” explains Yariv. “For me it was a huge thing, but then I listened to podcasts and stories. I learned. A pivot is really great — you assessed market fit and found this was a better idea. I was afraid we’d look indecisive, but in fact it’s the other way around.”
Finding the niche
Despite the pivot, the big vision remained the same. They were addressing a real and daunting problem; something they believed would affect everyone in the world.
“Machine Learning is changing everything,” explained Yaniv. “And we didn’t want to see it done carelessly. Our original product was to collaborate, automate, and validate the data inputs for machine learning models. But after the conversations we had, we just went with validate. When we focussed on the problem we saw the biggest pain point. But also the biggest amount of work. Are we going to work on something that has minor benefits? Or something that has a clear and important benefit, but takes more to build?” It’s no surprise they went with the latter.
Yariv likens their pivot to a refining of their focus. “We needed to reframe the issue and the solution. We moved towards a more niche and more pressing issue; one we’re destined to see more of in the future. So the pivot was also a ‘zooming in’ — it was focusing. We found a niche.”
badook’s advice when approaching a major shift in product development? Be married to the problem. Pivot to find the solution.
Why it all matters
The engineering mentality runs deep in these two. If you see something broken, you need to fix it. “We saw how test driven development and DevOps improved the software industry, but when it came to AI and Machine Learning validating data, the process was manual and difficult.” explained Yaniv. “The work is complex and messy and people shy away from it because it’s hard. There’s no clear solution for data issues so the inclination is to ignore it.”
By virtue of adopting machine learning, more and more organisations are getting burnt by compromised data sets. Yariv speaks to badook’s place in the market. “We’re ahead of the curve on this. The focus is education, and finding advocates in early adopters. But in time, we know it’ll be more mainstream.”
The co-founder and cultural dynamic
Trained in high tech, and from a country where entrepreneurship is engrained in the culture, this Israeli co-founding team found a little culture shock in moving to Australia. “The big culture shock was with enterprise, in how they behave and how decisions are made. It’s not just startups that are agile in Israel; it’s enterprise too.” explains Yaniv.
Yariv provides more colour. “It took me a while to recognise the rules of the game. The first thing? The pace is different. In Israel, we’re on the go all the time. I mean, we’re running from missiles back home! It’s a war field.” he says in jest. “The pace is born in you. Understandably, Australia is not the same.”
So how are these two Israeli engineers with very similar names different? Yaniv is quirky, tangential, sometimes chaotic, with both an endearing and fierce ‘Let’s do this!’ attitude. Yariv takes the chaos and finds the focus. He narrows in on the vision; is organised; is grounded. He’s on the receiving end of the heart attacks. While Yaniv is making an exploratory mess, Yariv is the one who exclaims “What did you do now?!”
At the heart of it, they complement each other. Yaniv spitballs ideas. Yariv will reflect them back to him, testing, and playing devil’s advocate. He’ll make him fight for it; to rationalise; to refine. And the result is stellar.
Start as fast as you can, and then go faster
The pace is built in their culture. In building their team, this culture continued to be fostered, almost unconsciously. “When you’re working hard and fast, and pivoting, and you’re asking potential employees to take a risk with you, we were fortunate to find like-minded engineers who went “Sure. Let’s do that!” says Yariv of their recruiting.
The team energy is not just about pace, it’s about exploration and fun. They’re not as structured as some might be, but they get shit done. In the Israeli startup ecosystem, the only thing anyone cares about is what you get done. At the end of the day, did you deliver or did you not?
What’s the big picture?
Bad data affects not just machines, but also humans. Humans being exploited by fake news is the same as corrupt data poisoning machine learning. This is a defining moment as a human race.
“We see it even now with COVID-19 conspiracy theories.” explains Yaniv. “We need to start treating facts and data in a more scientific way. We can’t afford to have more models taking a partial anecdotal piece of data as truth. Bad data is the biggest problem we have.”
He continues. “With Covid, the chaos we’ve seen is more about the intelligence and decision-making process than anything else. We’ve got to constantly question and verify what we’re being fed. We need to cultivate a knowledge of what is real and what is not. It’s more important now than ever.”
