The AI revolution: Adoption rhymes with interpretability

Artificial intelligence (AI) has taken the world by storm. Everyone is excited to use it, but perhaps few truly understand the complexities of machine learning (ML). However, that does not mean that non-technical people or companies should not be able to engage in meaningful conversations about the field or use AI solutions. Malik Boudiaf, a Machine Learning engineer, describes how Coactive’s team is educating its customers along the way by answering their sometimes very philosophical questions about the product. The AI revolution is here, and Coactive wants to bring companies along for the journey.  

Malik, now that you’ve been on the Coactive team for a little while, tell us… have we lived up to your expectations?

Coactive has absolutely exceeded my expectations! I was not originally sure how I would fit into the larger team while being based in Montreal, alongside Seby, while most everyone else is in San Jose, California. However, we definitely all operate as one team and it is because of that that I continue to be excited to be a part of the Coactive community. Also, aside from everyone being so incredibly welcoming, I have to say that I am surprised at how complementary everyone’s skills and backgrounds are.

Some people have worked with very optimized production codes with complex infrastructures, while others have more of a research background and explore all of the fresh trends happening in AI. It’s amazing to work with such a well-rounded team.

What about having a large mixture of backgrounds do you think could be so tumultuous?

I guess it wasn’t just the variety of backgrounds that was surprising, but more of how harmoniously everyone works. I have seen how differences can cause turbulence, but at Coactive all of our team members are genuinely focused on moving our company’s mission forward and doing so together. There is truly a culture of stopping and listening to the viewpoints and opinions of others. I was astonished when, despite having less experience on a topic, my colleagues really intently listened to what I had to offer and engaged my ideas. That simply does not happen everywhere.

Not only do we at Coactive work hard to communicate with one another efficiently and effectively, but we have a high level of trust that has arisen amongst us. People feel really responsible for both the solutions we develop, as well as the teammates that they develop with. I enjoy working with a team that is so focused on lifting one another up and finding the best ways to receive and transfer knowledge from each other.

How nice that where there could be conflict, Coactive is focused on harmony! Speaking of differing backgrounds, we’d love to learn more about you, Malik.

My whole life is a story of differing backgrounds! I was raised in a small town on the Mediterranean coast of France, but my parents immigrated there from Algeria. Of my siblings, I am the only one who has left France, as I now live in Canada with my fiancée. Ironically, my great uncle (Mohamed Boudiaf), prior to his assassination, was a primary political leader of the Algerian War of Independence - that led to Algeria resting its independence from France in 1962. I feel like I’ve always lived at the intersection of difference - which is part of the reason why I see Coactive as something special. It is so important to pay attention to and understand diverse perspectives, especially those that may be underrepresented. We do that here.

What a revolutionary lineage that beset an incredible perspective. Given your experiences, what draws you to machine learning?

To start, I have always been fascinated by the sciences in general. Ever since I was a small child, I have had a natural propension for areas like mathematics and scientific reasoning. Actually, my formal experience prior to completing my PhD was not computer science, but was in aerospace engineering. It’s funny that during the time I completed my Master’s degree in Aeronautics and Astrophysics at Stanford University, that I never met you (Danielle) or Cody. We were all there at the same time. But it’s especially interesting that Cody and I never crossed paths, because it was at Stanford that my burgeoning interest in computer science began.

How funny that the stars didn’t align in that way. How did your experience at Stanford overthrow your previous interests and peak a passion in computer science?

In some ways I feel like Stanford was the birthplace of AI. Whenever I had the chance to take classes outside of my required course load, I would consistently find myself learning from some of the leading people creating in the field of artificial intelligence. I just felt like it was going to be the biggest revolution to happen in the near future - and in some ways already is happening in the present. The field of computer science really aligned with everything that I love: practical and influential applications, coding, actual production, etc. It’s where the robot hits the road.

So, I decided then that AI and computer vision is where I wanted to be. Not long after finishing my master’s degree, I was pursuing a PhD in the field in Canada.  

What an academic regime change! After your PhD, how did you decide that Coactive was the next stage for you?

As I neared the completion of my doctorate, I honestly had a lot of different options in academia. However, one of my passions is to build real things for the real world. To me, where all of this research meets real world needs is where things get really exciting, because you have the opportunity to solve problems that other people are facing.

It was serendipitous, because just as I was deciding whether I wanted to go into industry, I found Coactive online and instantly could see the many different applications. One in particular was a solution for my fiancée’s line of work as a surgeon. I saw that there could be a very easy fix to an issue she is having with showcasing videos from her lab - all through the prowess of machine learning. The conversation I had with her really motivated me to go to a company where they’re democratizing AI and making it useful for non-technical folks. So I applied and the rest is history.

In what ways do you think Coactive is rebelliously leading the way with regard to the field of AI?

One way we’re pushing the envelope is through multimodal understanding, i.e. processing information from multiple sources simultaneously (visual, textual, and audio cues). We’re literally developing a whole new generation of automation tools, and already changing the game on several crucial problems. Content moderation is one of them. For example, content can be inappropriate due to purely graphics, or because it contains written or verbal slurs/hate speech, etc. Automation is absolutely crucial to cope with the scale of the problem, and to alleviate the trauma this type of work can cause to moderators. At the moment, moderators need to manually eliminate the most disturbing material online. Coactive’s automation tools help moderators gain orders of magnitude in productivity and efficiency.

Further, rather than just giving a solution to a business, we really prioritize educating our customers along the way. I have worked as a tutor and teacher much of my life and I feel like those effective communication skills are called upon as a facet of this job. We work with clients to teach them how and why our solutions are the best choice for them. Rather than just sell people products (which we could do), at Coactive we work to educate them about all of the incredible things we can do with AI.

[Malik running Oxford’s famous Town and Gown 10k race!]

Have you had any interesting conversations with technical or non-technical customers that have shaken the table when it comes to customer education?

Absolutely, and it’s been some of the best parts of the job! I feel as though we’re all on a learning journey together sometimes, as our conversations can get quite philosophical - not just technical. For instance, we presented a demo to a large media company recently where our discussions turned to interpretability. They were really excited about what we produced, but also had questions like:

- How did we come up with these results?

- How do we guide your system of ML in a way that allows us to productively use this solution, rather than having the ML guide us?

- How can we use this in a way that doesn’t lead to an eventual replacement of our employees?

Honestly, it took me off guard the first time it happened, but in such a good way. Companies don’t want black box systems. Instead, we work with them to help them understand the science behind what we do so that our solutions aren’t just spitting out “truths,” but instead offer interpretability. People are really excited about AI and the results they can get from the industry, but they are even more curious how they can be given the keys to understanding how the platforms and models perceive the world. At Coactive, we engage in those conversations because we respect the people behind the companies we are selling to and we know that having those higher-level conversations can be important to the successful use of our systems.

That is such an authentic way to engage with the burning questions that many people have around AI. What are you most excited to be working on currently?

I mentioned before that I revel in solving real world problems that every day people are having. I really mean that. Right now, I’m working across many projects that are fun, but what I love the most is the opportunity to quantify the value that we’re bringing to companies. Showing people with actual numbers how we’re able to help them achieve the goals of their workflows more productively and with greater efficiency is an amazing part of what I get to do at Coactive. Designing benchmarks or “golden data” that consistently highlights how much effort we’re able to save individuals and teams, on multiple fronts, is in part exactly what I hoped I would be doing when I left academia for industry.

Sounds like your professional goals and the aims of Coactive are aligned. How perfect! To close, tell us what brings you joy outside of work?

While it may seem like the opposite of the career choice I’ve made, I really love woodworking. I genuinely enjoy building with my hands. I originally started creating small toys and toddler tools for my nieces and nephews. Now, I’ve graduated to building tables and refining my skills through taking advanced classes. Machine learning and woodworking are a great complement to one another to me, despite their apparent differences. They’re both about exercising your mind and/or body to create something incredible.

[Malik's first original creation: a toy airplane!]

We can definitely agree on that. Thank you for your time, Malik!

Coactive AI is an industry-leading machine learning platform that enables businesses of all sizes to analyze their unstructured image data in minutes. Our interface is clean, intuitive, and user-friendly, and our platform is blisteringly fast. Engineers, if you want to help us build the future, check out our job openings. Businesses, if you want to harness the power of machine learning like never before, request a product demo today.