Data should trump AI hype, Lazada co-founder tells CEOs

Magnus Ekbom, co-founder of ecommerce platform Lazada, isn’t fond of questions packed with industry buzzwords. At the Tech in Asia Conference in Jakarta last week, he made his opinions about jargony words clear.

When asked what he thinks of scale-ups, he says the word sends people’s thinking “down the wrong path.” When asked if his firm is a product company, a service company, or a consultancy, he pauses for a bit, then says, “I hate labels.”

This isn’t Ekbom being cantankerous. Being one of Lazada’s founding members, perhaps generalities just don’t suit him.

But Ekbom – who has just launched a new firm called SEAI – thinks getting the data right is critical for the future of AI.

Image credit: Timmy Loen

The new company – launched in June along with Ashwath Ramesh, ex-Lazada COO for Asia Pacific – is focused on customized software that uses machine learning to “solve real-world problems” for clients.

SEAI is building a very similar suite of software services to what Salesforce offers, barring messaging platform Slack. Headquartered in Singapore, the firm hopes to see South Korea, Japan, Taiwan, Australia, and possibly even Europe as “core markets” in a few years’ time.

“Successful machine learning starts with strong data foundations,” Ekbom says.

He adds that SEAI spends an outsized amount of time building these data foundations, which are far more important than the models.

“Everyone wants to talk about the models,” Ekbom points out. “The ballgame is the data – how rich, how clean, how well-structured your data is.” Once that’s established, different models can be applied to train and fine-tune the AI.

Good data doesn’t come easy, he notes. In many companies, data sits in multiple databases, some of it not even digitalized. Startups need to be willing to do the hard work of making that data available in a format that AI models can work with.

That said, he points out that such concerns may be less prevalent in Southeast Asia than in other markets because a generation of businesses in the region, particularly SMEs, are digitally native. But there are still businesses struggling to get the right data for their AI models.

See also: Lazada co-founder enters AI arena with Salesforce-like product

Data, not AI, should be the CEO’s focus

Ekbom points to the healthcare sector, which he calls “the single largest generator of data worldwide.” About 30% of the world’s data volume is generated within healthcare, but he estimates up to 80% of that never reaches a database.

What’s behind the AI boom

Garbage in, garbage out

Like building the railroad