Oracle has set up an AI Centre of Excellence in Singapore which prioritises building working solutions rather than creating demonstrations, so as to go past the hype that had beset many early AI efforts in the past 18 months.
Opened this week, the facility welcomes multinational corporations, startups and SMEs alike, promising consistent service standards regardless of client size or market position.
Customers will be able to choose the AI model they want for their application. There will be a dedicated team running the centre and a number of partners to help businesses on their AI journey. Partners include ST Engineering, NCS, Accenture, Deloitte and NTUC LearningHub.
Dr Tan See Leng, Singapore’s Minister for Manpower and Second Minister for Trade and Industry, announced this at the one-day Oracle World Singapore on March 13.
“The new Oracle AI Centre for Excellence is a good example of how businesses can take the lead to leverage new technology effectively,” he said.
The Singapore facility will provide the knowledge for enterprises to experiment and test early-stage AI innovations in secure cloud environments and enable organisations to quickly adopt new predictive, generative and agentic AI features in business processes in departments like finance, supply chain and marketing.
It will also provide organisations access to training sessions and certifications on the latest cloud and AI technologies, led by Oracle University and business partners. Oracle hopes to train 10,000 students and professionals in Singapore with the latest digital technologies skills by 2027.
“We’re not creating a cool meeting room with screens,” said Garrett Ilg, executive vice-president of Oracle Japan and Asia-Pacific. “We’re building actual solutions.”
The latest initiative aligns Oracle with Singapore government’s push to get global tech firms to establish AI facilities in the Republic.
Oracle, which had revenues of US$53 billion in the 2024 financial year, also announced other initiatives at its event, including the availability of its database services on Microsoft Azure data centres, as well as new anti-fraud AI capabilities.
The anti-fraud agentic AI capabilities are expected to help financial institutions combat financial crime. These intelligent tools automate complex investigative processes, efficiently uncovering sophisticated patterns that might otherwise go undetected.
The AI-powered system is said to enhance both quality and efficiency in reviewing suspicious activities by generating comprehensive analytical narratives that complement human investigators’ work. This reduces the burden of manual tasks, allowing investigators to dedicate their expertise to high-priority leads.
In September last year, Oracle said it would ship more than 50 AI agents, as part of its Fusion cloud business applications suite, in finance, human resource, supply chain management, quality control, sales, and customer service.
Sunil Wahi, vice-president for solution engineering applications at Oracle Asia-Pacific, said agentic AI are autonomous work packages, intelligently automate and streamline business operations with minimal human intervention.
What sets apart these AI agents is their contextual awareness, he noted. They continuously scan the context and domain they are operating in to identify and extract relevant information.
This deep contextual understanding enables them to adapt their processes in real-time, making informed decisions that consider the full operational landscape.
This results in business efficiency and increased productivity, said Wahi, on the sidelines of the conference. It also reduces the “number of engagement clicks”, benefitting users and allowing them to focus on decision making, he added.
While the company’s agentic AI solutions currently work in vertical domains like finance, sales and marketing, the aim is to extend the such systems to operate horizontally across all business functions and industries.
This cross-domain approach aims to create interconnected intelligent systems that can share contextual insights and coordinate processes throughout entire organisations. This way, end-to-end business processes rather than isolated departmental tasks can be addressed, said Wahi.
He stressed that effective agentic AI systems rely on seamless data integration. Organisations should implement strategies to consolidate their data from diverse sources into a single unified platform, enabling AI to efficiently query and synthesise information for optimal performance, he added.
On hallucinations, he said that human oversight is still needed. Hallucinations can emerge, despite rigorous data curation. While AI agents excel at “reasoning” through complex scenarios, their logical processes can occasionally produce flawed conclusions or generate inaccurate information, he stressed.
