This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy

AI enables our analysts to go up a level
07

“Five steps ahead”

Amien Krisna, CTO of Dataxet

Gavin Allen: Your company specializes in media monitoring and analysis. Does AI pose a threat to your industry?

Amien Krisna: We started out as a traditional media monitoring business. Now, AI is actually helping us fine-tune a lot of processes and get reports out faster.

As for the research, AI enables our analysts to go up a level. It does all the basic work, to which our human analysts add a creative, intuitive layer. So clients actually get better stuff—and they get it faster—because all the mundane work is done by AI.

Gavin Allen: So instead of being a threat, AI actually allows you to enhance your service?

Amien Krisna: Correct. It enables scale and allows our team to be more creative. Clients get better insights and analysis.

Gavin Allen: Now, you're working closely with Huawei Cloud. What are the solutions and services that they're offering that are helping you in your industry?

Amien Krisna: We've been with Huawei Cloud since early 2020. Currently, we have a platform where almost everything is on the cloud. This helps us scale our services faster than we could do if we were still using basic co-location hardware.

Gavin Allen: You're halfway through building a serverless social media analysis platform—one that allocates resources on demand and scales up quickly to meet spikes in demand. Why was that important for you? And what challenges has it presented?

Amien Krisna: We did it because of scale and cost. We have a Natural Language Processing model that analyzes languages across southeast Asia. So, Thai, Bahasa Indonesia, Bahasa Melayu [Malaysian]. We actually have to hit those services quite frequently, because every time a post comes in, we have to tag and automate that data so we can make sense of it for our clients. Serverless access helps us do that faster. Every time there's a new service, we don't have to build a new virtual machine, install it, and configure it. With the serverless, it can go straight from day one.

So I think that's basically the benefit. The challenge is fine-tuning our current software model to actually fit those services. Luckily, we've done it since 2020, so it wasn't that painful migrating there. But there were still training levels, getting the right people in, enrolling them in significant training models.

But Huawei has been very good at that. They have these training programs so that, when we hire somewhere, the person can just go straight into a learning program.

Gavin Allen: So is it a constant adaptation, a constant flexibility that's required for all industries now?

Amien Krisna: It is. The technology is moving so fast, and AI is moving faster. I mean, every year, there's something different. All the models we have to build, all the development—we have to be there. With that kind of setup, it enables us to scale, build, and deploy a lot faster.

Gavin Allen: What’s next for your technological partnership with Huawei?

Amien Krisna: We have a few advanced AI models and NLP models in the works; they’re currently in our R&D department. We're looking to explore what kind of infrastructure Huawei can support and how we can take advantage of the GPUs to make our services faster for clients.

There's a lot of things. I want to see how we can just get a bit faster. Because it's a constant, I wouldn’t say battle, but a challenge. Every year, there’s something new. Our customers want something different. So we have to push the envelope every year.

Gavin Allen: If you stand still, you go backwards.

Amien Krisna: Exactly. If there's a technology or service, and we don’t do it, somebody else will. So, we have to constantly be five steps ahead. That's where the challenge arises.

Contact us! transform@huawei.com