Cloud Hosting

What do Hong Kongs business giants have in common? A diverse portfolio of businesses, long history, and large operations. These characteristics also make them less agile to change. But many remain successful because of their ability to disrupt and innovate.

Against the backdrop of the rise of AI, many of them are going through the next phase of transformation. Enterprises are rethinking their core data architectures with data lakehouse platforms as they prepare to ride the oncoming AI wave.

A data lakehouse is the latest data management architecture that combines the flexibility and scalability ofdata lakesin storing structured and unstructured data with the data management and transactions of data warehouses, enabling BI and machine learning (ML) on all data. More about data lakehouse can be found here.

Shaking from the data core

One of them is Li & Fung, a leading global supply chain player. The 117-year-old company is rich with history and talents but also complex. Leo Liu, the chief digital officer of LFX, the digital arm of Li & Fung, recently shared the century-old companys data transformation journey at the recent Data + AI World Tour conference in Hong Kong organized by Databricks.

With over 100 years of supply chain knowledge and experience, LFXs mission is to create and invest in digital ventures that will transform the supply chain and retail industries, said Liu. One key initiative is to modernize our legacy data platform.

He explained the company once acquired 50 different businesses within a year, creating an extremely complicated environment with multiple technology stacks. On top of that, the organization relied on a 10-year-old on-premises data warehouse, slowing down its big data and AI strategy for innovation. AI cannot work without [a modernized] data platform, he said.

Shaking down a data platform from its core for a business giant is not easy. But Liu took the challenge to a new level by looking to complete the data architecture transformation within five months. This includes configuring, setting up a new cloud, data platform and data pipelines, migrating the data and dashboards, training, and launching.

Im still here, so you know we did it, Liu said. The success includes producing three quick wins to demonstrate the value of the new platform, all within the year.

On top of his dedicated and motivated team, Liu said the success was attributed to three primary criteria of the new data platform: open standards; multi-cloud support; fully integrated for data engineers, scientists, and business users. Working with Databricks in building a data lakehouse platform, Liu said the company could now achieve all three criteria and focus on its AI innovation.

We can now develop dashboards and reports within 24 hours; before, it took weeks and months, he said. This year, we are moving forward with our AI strategy.

Beyond legacy

Dealing with legacy is even more challenging for highly-regulated payment players like HSBC and Octopus Card. They are achieving better data governance and predictive modeling while riding on a data lakehouse platform.

As business needs to evolve, there is a growing need for better data analytics and robust data governance to ensure that data provides value and supports our business strategy, said Thomas Qian, wholesale chief data science architect & analytical platform lead at HSBC.

Qian noted one example is the tracking and analysis of users behavior at PayMe, HSBCs mobile payment service. He said the insights on customers usage patterns contributed to the launch of PayMe for Business, a service for merchants to collect payment.

By working with Databricks, we can scale data analytics and machine learning to enable customer-centric use cases, including personalization, recommendations, and fraud detection, he added.

Data governance struggle

Meanwhile, at Octopus Card, data privacy and governance have been top priorities.

We have a very tight data governance policy to protect customer data because we literally hold the data of all Hong Kong people, said Tony Chan, senior data science manager at Octopus Card.

Chan shared the challenge of battling through the stringent governance process to access data. But his team is exploring the use of a data lakehouse platform for easier data governance and more scalable prediction analysis.

He added that they wanted to move from rule-based analysis to AI modeling to detect merchant churn rates. The data lakehouse platform allows his team to scale the analysis on tens of thousands of merchants and predict their churn rate, helping the sales team to prioritize the renewal process.

We hope to slowly transform the users' and senior managements expectations on AI and promote more AI applications, Chan added.

Speedy rebound as borders reopen

The data lakehouse platform is also transforming customer experiences. Swire Properties, a real-estate unit of another Hong Kong-based conglomerate, has recently taken advantage of a data lakehouse platform to drive precision marketing.

I first joined the company as customer relationship management, nothing technical. But I quickly realized that I could do nothing about quality customer engagement without quality data. So, I took the initiative to formulate a data strategy, said Veronica Ho, head of data analytics & insights at Swire Properties.

Part of the strategy was consolidating more than one million data points from 30 different data sources across four business pillars: shopping malls, offices, residential, and hotels. This consolidated data platform became the foundation for developing a predictive model supporting the companys speedy rebound from the post-COVID-19 border reopening.

Ho said by understanding the customers with multi-faceted segmentations, the company can develop hyper-personalized and precision marketing campaigns, like tailored birthday surprises. These precision marketing campaigns allowed the company to reach and engage seven times more members.

For members across the border, data-driven marketing also allowed the team to identify and develop personalized offers to 60% of high-potential members who went dormant during the pandemic.

Like Swire Properties, Octopus Card, HSBC, and Li & Fung, unifying the data platform to drive data integrity and governance is only the first step towards realizing their AI strategies. More data-forward business giants are harnessing the value of their data and applying AI through the lakehouse platform to transform into business legends.

As pioneers of the data lakehouse, we are passionate about making data and AI accessible to everyone, said Jia Woei Ling, managing director for North Asia at Databricks.

Sheila Lam is the contributing editor of CDOTrends. Covering IT for 20 years as a journalist, she has witnessed the emergence, hype, and maturity of different technologies but is always excited about what's next. You can reach her at[emailprotected].

Image credit: iStockphoto/Orla

Read more here:

Perfecting the Data Balancing Act - CDOTrends