InfuseAI, a MLOps startup in Taiwan, has joined the AI Infrastructure Alliance (AIIA) as one of its core members. It is also the first Asian company in the alliance. By joining AIIA, Taiwan has also become a force in the global effort to democratize AI.
Formed in February 2021, the AIIA is a consortium of more than 30 leading AI startups seeking to tie together a complex web of existing AI technologies into a single Canonical Stack. Other core founding members include Pachyderm, Seldon, Determined AI, Algorithmia, Tecton. Together, they have raised over US$200M in collective venture capital funding.
Solving an infrastructural problem
As AI and Machine Learning (ML) underpin the wave of digital transformation unfolding across industries, large companies with cutting-edge capabilities – especially Big Tech like Microsoft, Google and Amazon – continue to dominate AI development and render it hard for smaller companies to access ML’s powerful potential and define AI on their own terms.
For AIIA, the alliance’s role is to build a truly common infrastructure for AI through establishing engineering standards and integration points – a Canonical Stack for AI – as opposed to the infrastructure currently built by the Big Tech with their vested interests.
A bridge in digital transformation
InfuseAI builds PrimeHub, an open-source MLOps platform, that provides modules for smaller enterprises to seamlessly plug in AI models and scale up their AI deployments. Currently, standardized toolkit and solution are absent in the AI/MLOps field. InfuseAI’s participation will help AIIA to build the Canonical Stack in the AL/MLOps field.
True to the founding vision of AIIA, the founders of InfuseAI aim to be a bridge in digital transformation, providing a foundational service for enterprises to swiftly build up ML environments.
In doing so, it addresses a tech gap that is a common debilitating factor faced by many enterprises aspiring to deploy AI.
At the current stage, InfuseAI targets enterprises that already possess the necessary hardware and data to build an ML environment. In the future, InfuseAI also looks to cooperate with hardware providers to service enterprises that seek AI deployment without the necessary hardware.
For AIIA’s vision of a trickle-down AI, InfuseAI’s inclusion sets another milestone in Taiwan’s contribution to AI governance: an issue of growing importance when cutting-edge AI development is no longer a mere issue of technological progress, but a political one as well.
Especially when Big Tech, concentrated in the US and China, have inevitably become their national governments’ proxies to shape political governance.