How do you turn a great idea into a product that delivers real value? Throughout the Future Says series, it’s a question that’s been highlighted time and again by industry experts employed in artificial intelligence (AI), machine learning (ML), and data science. But for Sahar Asadi, our guest for episode five in this second season of Future Says, that issue sits at the heart of her work day in and day out. As AI research lead for King, a leading interactive entertainment and mobile gaming company, Asadi is responsible for operationalizing blue sky thinking. What’s more, she’s doing it in an industry that’s exerting a powerful influence over the direction of the tech industry at large.

Gaming is already big business, and it’s only getting bigger. By 2025, the industry will be worth an estimated $250 billion annually. Within this powerhouse sector, King is a well-established name. Responsible for numerous global gaming hits, including Candy Crush, the company currently attracts a staggering 255 million active players per month. But the wider impact of the gaming sector isn’t simply a product of its scale. From the outset, gaming has been a testbed for AI and ML innovation. Novel applications in areas like content automation, personalization, and the creation of more realistic gaming environments routinely provide the basis for solutions in other industries.

Asadi is on the front line of this groundbreaking work. As a result, she has invaluable insight on how to translate research into practical, viable AI and ML products. In doing so, she covers both the attitudes and the environment needed to foster the right blend of creativity and realism, as well as the growing significance of technologies like accessible ML Ops tools.

“In the discovery phase, we basically go crazy,” explains Asadi. ”We try out ideas, we look at previous work, we look at data and opportunities. After this initial assessment, we look at estimates of how long this research will take, the potential impact it can bring, and try to pitch that as an idea to get prioritization on.”   

Asadi stresses the key role the democratization of ML Ops platforms is playing. Consequently, she believes it’s time to move away from the idea of data scientists as “the Swiss Army Knife of solving everything.” With more intuitive ML Ops platforms, a wider range of employees can contribute to the process. Data scientists now, therefore, have more freedom to apply their expertise and experience where it’ll be most impactful.

Previous Future Says interviewees, drawn from diverse sectors like manufacturing, finance and consultancy, have all highlighted the challenges of operationalizing AI and ML. But another common thread that runs through this series is the degree to which the answers are universal. So, while Asadi dedicates her energy to the surreal task of bringing alternative universes to life, her parting thoughts will undoubtedly resonate with previous and upcoming Future Says guests: “Be patient, be realistic, build a culture, and, for researchers, understand the need before jumping into the state-of-the-art and putting that in production.”

To watch the full interview with Sahar Asadi and all the preceding episodes of Future Says, click here.