Across Malaysia, enterprises are feeling the urgency to embrace artificial intelligence (AI) and revamp their operational frameworks. However, the journey towards AI integration necessitates a departure from outdated legacy systems, steering towards environments that promise both cost efficiency and scalability.
During the roundtable discussion titled “Modernising Databases and Scaling in the AI Era,” hosted by Jicara Media in collaboration with EDB Postgres AI, senior IT leaders gathered to share insights on the prevalent hurdles in AI deployment. A significant focus was placed on the need to curtail unnecessary expenditures associated with legacy systems and to pivot towards investments that foster innovation.
Leaving legacy behind
Legacy infrastructure has become a common anchor, restraining enterprises from fully realizing their potential. One organization, for instance, embarked on a modernization journey in 2021, yet continues to grapple with several applications still tethered to legacy systems.
“We have adopted a hybrid cloud approach. We’re utilizing SQL and various databases across more than 100 applications. Given our long-standing presence in the industry, managing a vast amount of data is quite a challenge, especially with numerous subsidiaries and departments,” shared a senior manager from the organization.
Similarly, UEM Edgenta, an asset management firm, has faced the dilemma of outdated applications that have remained untouched for decades. Chief Digital Officer Chua Yong Howe expressed the need for modernization to extract insights from these legacy systems. “Our goal is to ensure that as our team accesses the database, they also acquire institutional knowledge. However, my immediate concern is maintaining application availability during peak usage periods,” he remarked.
Chua highlighted their reliance on a major enterprise database system, complemented by a cloud-based data analytics platform, alongside traditional relational database systems.
AI pressure
The push for AI integration is palpable across various industries, driven by both management directives and customer expectations. Tze Phei Tee, CIO of energy solutions provider Wasco Berhad, recounted a conversation with their CEO regarding the exploration of agentic AI. “He urged me to investigate its potential in HR and finance, particularly in back-office functions. We are currently in the process of identifying suitable use cases for agentic AI, while also assessing our readiness for the necessary transformations,” he noted.
Tze pointed out that employee willingness to undergo upskilling could pose a challenge in the AI transformation journey. “Not all employees are eager to enhance their skills; many are comfortable with their established routines. This is a common challenge that organizations must address,” he added.
Seah Boon Chong, Head of Sales Engineering for APJ at EDB, emphasized that not every challenge necessitates an AI solution. Drawing from customer experiences, he suggested that starting with smaller use cases often leads to quick wins that can be scaled later. “It’s crucial to evaluate whether AI is genuinely required. In some cases, analytics might suffice, with AI being the next logical step,” Seah explained.
Reliable support
As enterprises increasingly adopt open-source database systems like Postgres, ensuring reliable support becomes paramount. Timely intervention is essential for addressing minor issues, such as bug fixes, as well as major incidents like service outages. EDB Postgres AI positions itself to assist organizations in mitigating vendor lock-in while offering enterprise-grade support for Postgres deployments.
“Open-source Postgres is extensive, but for mission-critical applications, downtime is not an option. We provide enterprise tooling and support to ensure the security and availability of your applications,” stated Ang Lee Yen, Asia Sales MD at EDB.
Seah further elaborated on the complexities of data sovereignty for enterprises operating in both on-premises and cloud environments. Organizations often face challenges in managing siloed data, vendor lock-in, and intricate workload orchestration while adhering to regulatory requirements. “When discussing sovereignty, we must consider data extraction, protection, and control within local data centers,” he explained.
EDB Postgres AI offers a sovereign data and AI platform that merges the security and control of bare metal with the agility of the cloud. This platform allows enterprises to run transactional, analytical, and AI workloads while ensuring that highly regulated data remains on-premises to comply with legal standards. Simultaneously, other workloads can leverage the cloud’s flexibility and scalability, all managed through a consistent experience across deployments.
Seah concluded by highlighting the importance of reducing infrastructure costs to free up resources for higher-value initiatives. “If you do not minimize the expenses associated with your existing infrastructure, which consumes substantial resources, your capacity to invest in new initiatives will be limited,” he remarked.
