Infrastructure AI Launches Agentic Hub 1.0 to Bring Persistent AI Intelligence to Buildings, Factories and Smart Cities
KUALA LUMPUR: Infrastructure AI has introduced Agentic Hub 1.0, a new artificial intelligence platform designed to embed persistent AI intelligence directly into physical infrastructure, marking a significant step towards autonomous management of buildings, industrial facilities and public assets.
Unveiled on Tuesday, the platform aims to move beyond conventional AI systems that perform isolated tasks by creating AI entities capable of continuously learning, remembering past events and making increasingly informed operational decisions throughout the lifespan of physical infrastructure.
According to the company, Agentic Hub 1.0 is built to operate across a broad range of environments, including commercial buildings, manufacturing plants, utility networks, airports, hospitals, university campuses and transportation systems.
Unlike traditional AI applications that are activated only when a specific task is requested, the new platform is designed to function as a permanent digital presence residing alongside physical assets.
Infrastructure AI said the platform combines two complementary forms of artificial intelligence within a secure edge-computing environment.
Neural network agents continuously analyse operational data by detecting anomalies, recognising patterns, predicting equipment behaviour and monitoring system performance. Working alongside them are large language model (LLM)-based agents responsible for reasoning, planning, coordinating workflows, consulting digital twins and supporting operational decision-making.
The company said integrating both AI approaches enables infrastructure to evolve from being manually managed to becoming increasingly self-managing through continuous learning and accumulated operational experience.
At the centre of the architecture are secure “Agentic Containers” hosted inside Infrastructure AI’s proprietary Intelligence Interface Module, allowing AI agents to maintain persistent identity, operational memory, contextual awareness, governance controls and digital twin integration while remaining close to the physical systems they supervise.
Infrastructure AI describes this as transforming conventional edge computing devices into what it calls “Living Infrastructure Devices”, where AI continuously observes equipment, learns from real-world operations and improves its own understanding over time.
The platform has also been designed with governance and security as core principles.
Each AI entity is envisioned to possess a persistent digital identity, defined operating permissions, governance policies and auditable activity records, supported by technologies including blockchain-based identity management, digital twins and swarm intelligence frameworks.
Co-Founder and Executive Chairman Dilip Rahulan said the combination of predictive neural networks and reasoning-based LLM agents represents a new architectural model for infrastructure intelligence.
He added that building trust would be essential before autonomous AI systems are widely deployed across critical infrastructure such as utilities, airports, factories, hospitals and cities.
Meanwhile, Co-Founder Glen Allmendinger said the greatest value of persistent AI lies in its ability to continuously accumulate operational knowledge rather than simply completing isolated software tasks.
As intelligent systems collaborate across multiple assets, he believes entirely new network effects and economic opportunities could emerge through shared knowledge and long-term operational learning.
Infrastructure AI ultimately envisions Agentic Hub evolving into a globally connected infrastructure ecosystem in which millions of intelligent AI entities cooperate across buildings, transportation networks and public infrastructure, creating what the company describes as an emerging “Intelligence Economy” for the physical world.
