Tiger Research releases IoTeX analysis report… “Shift to an AI platform, build real-world data infrastructure”

Tiger Research, a Web3-focused research firm, said on the 20th that it has released an analytical report on the blockchain platform “IoTeX.”

The report said IoTeX is transitioning, on the back of the DePIN infrastructure it has built over the past eight years, into an integrated infrastructure that supplies real-world data to artificial intelligence (AI).

According to the report, while AI is delivering results across most areas, it fails to function properly the moment it depends on fragmented, unverifiable external data. Take autonomous driving as an example: it can work with only onboard sensors, but completeness improves markedly when thousands of external data sources—such as traffic signals, crosswalk sensors and weather stations—are connected in real time. The problem is that these data are operated by different organizations, come in disparate formats and lack even a common standard for verifying provenance. The report said IoTeX has built an integrated infrastructure over eight years to close this gap.

According to the report, IoTeX converts signals from the physical world into data that AI can use immediately through a three-layer AI technology stack of verification, indexing and perception. The first verification layer, “ioID,” assigns blockchain-based digital IDs to physical devices such as sensors and robots, ensuring data provenance and integrity. The second indexing layer, “Quicksilver,” collects sensor data scattered across hundreds of networks, processes it into a machine-readable format and delivers it. The third perception layer, “Realms,” adds industry-specific contextual knowledge on top of processed data, enabling AI to make judgments beyond simple figures.

The structural monetization challenge typical of middleware was addressed with the commercial product “Trio.” Trio is a multimodal, stream-based agentic AI product that analyzes video streams in real time. When connected to a live video feed, AI analyzes what is happening on screen and returns answers in natural language. Users can receive immediate assessments in response to questions about the footage without having to monitor CCTV directly. The report assessed Trio as strategically meaningful in that it moves beyond low middleware fees to generate tangible revenue.

There is also a push to expand the ecosystem. According to the report, in September 2025 IoTeX launched the “Real-World AI Foundry,” a platform where multiple companies pool infrastructure, data and computing resources to jointly develop AI models. Participants include Vodafone and Filecoin, and IoTeX takes on an intermediary layer that verifies and connects resources. The goal is to develop “Real-World Models (RWMs)” that train AI on data generated in real time from sensors and devices.

Yoon Seung-sik, head of research at Tiger Research and the report’s author, said, “IoTeX is transitioning into an AI platform based on the DePIN infrastructure it has built over eight years,” adding, “Trio’s enterprise customer contract performance and the Foundry’s level of model commercialization will be at the core of the 2026 investment thesis.”