Red Hat opens Singapore AI centre to take customers beyond pilots

Red Hat's new AI Centre of Excellence (CoE) in Singapore addresses a common challenge in enterprise AI adoption: moving projects beyond initial proof-of-concept (PoC) stages into full production. Daniel Aw, Red Hat's Asia-Pacific vice-president, stated the CoE will station engineers locally to reduce the feedback loop to headquarters and ensure customer requirements in the region are met. This includes integrating software not yet natively supported by Red Hat, potentially from local startups, provided it aligns with Red Hat’s platform strategy.

A key focus for the CoE is enabling what Red Hat describes as "sovereign AI capabilities." While many organisations currently consume AI by sending requests to frontier models through application programming interfaces (APIs), Red Hat's approach aims to transform them into "token providers." This means helping customers host their own AI models, particularly open-source models with open weights, within their own environments. This shift allows organisations to maintain greater control over their AI operations and data.

The CoE, initially staffed with around 18 people including engineers, architects, and project managers, plans to build reusable software architectures for common industry problems and contribute to open-source projects. This effort targets enterprise requirements such as system deployment and runtime optimisation. Yueh Chinn Seah of Digital Industry Singapore noted that the CoE's emphasis on production readiness distinguishes it from other AI centres, aligning with Singapore's National AI Strategy to expand its pool of AI practitioners.