From the Swedish Ultimate team to the cutting edge of machine learning: What Coactive is bringing to businesses

Jacob Greenberg is an experienced full stack engineer, who knows first-hand the challenges businesses face in operationalizing AI tools. He’s also a semi-professional Ultimate player, playing on the Swedish national frisbee team beginning in 2012. No matter the domain, Jacob is ready to tackle any challenge.

In this blog you’ll hear Jacob discuss how the challenges other companies are facing fuel him to build solutions for the sector.

Jacob, tell us about your rise to Ultimate fame.

It all happened in quite a short period. I started playing during high school, then joined a club team in Stockholm in 2010 before making the national team two years later. By then I was used to playing across Europe, but I joined the national team at the deep end - the first tournament I represented Sweden in was the 2012 World Championships in Japan. We narrowly lost the semis on a tie-break point! I was quite close to having a World Championships medal…

What would motivate you to leave Sweden after such a successful run?

No, it was actually a perfect trifecta of pull factors. My dad is from the U.S., so I already had family in the Bay Area. I also knew I would easily transition in the job market as a software engineer who was moving to Silicon Valley. But my main motivation was that I wanted to give the U.S. semi-pro league a shot. The Ultimate scene is a lot bigger over here and I wanted a new challenge, so I relocated. In my first two weeks here I got four job offers and I made the cut for two semi-professional frisbee teams, so the flight was worth it!

To employers, you’re clearly quite the catch. What’s your professional background?

My master’s research was about getting autonomous drones to fly in areas without GPS, which meant the drones were reliant on cameras and computer vision. After that I took a break from computer vision for a while and focused on web development.

When I moved to the US in 2018, I joined an AI retail platform called Stockwell as a full stack developer. During my tenure there, I became head of engineering, leading roughly a dozen engineers. The platform we worked on allowed anyone with a space – e.g. a college, a hotel, a mall – to place a pre-built cabinet that functioned like a vending machine and run an automated store from it. Kind of like a customizable, mini Amazon Go store. This naturally involved a lot of consumer-facing visual data and computer vision modeling.

Covid forced Stockwell out of business, and I decided to take a break from image and video data. I joined a gamified wine startup, and took on some new technical challenges focusing on the ecommerce side. Then I was introduced to Coactive AI via a mutual friend who also knew Steph. She explained that this new start up, Coactive AI, was solving a critical visual data problem that we had issues with at Stockwell.

So you threw your hat in the ring (very accurately) and joined Coactive AI. Can you elaborate on the pull factors for you?

If a business can’t efficiently process its critical visual data, it will never achieve its potential, no matter how good the underlying idea is. I’ve worked in AI companies that could have benefited from more efficient ways of processing images and video, because there was simply no off the shelf solution for them.

For companies using ML/AI for visual data, a large amount of their efforts are spent moving the data around into useful formats, and building the underlying infrastructure to do it. That leaves very little time to focus on the actual value proposition of their ML product.

A platform like Coactive would've strengthened the resources and elevated the capacity of my previous startups to innovate. I kind of wish I’d had it a few years back.

What problem are you currently working on?

Right now, I’m focusing on trust and safety. Inappropriate content is a big issue for companies dealing with user generated content. As a result, many companies have to invest a lot of time and energy in detecting and filtering it. I’ve been helping build this solution into Coactive’s offering. By automating the removal of offensive material we can help our clients ensure a safe environment for all of their users – whether employees or the public.

Is there anything that’s stood out for you, since you joined the team?

We’re working on such a wide variety of problems, which is a lot of fun. From front end CSS to backend infrastructure, we’re working with millions and millions of images so we deal with really interesting challenges. I’ve been excited by the opportunity to re-focus on developing and optimizing engineering solutions for large volumes of data to keep efficiency high and prices low.

One more thing that stands out for me about Coactive though is the culture – we have such a good collaborative culture in the company, everyone's very helpful.

Wow, what a perfect closing. Thanks so much for spinning discs with us, Jacob!

Culture is our number one product, and it was our first ever release. We pride ourselves on having a truly collaborative and ever-learning team. We’re also the industry leaders for analyzing unstructured image data. Are you ready to join the next revolution in machine learning? Check out our job openings.