Attentix AI
Attentix AI has built a dedicated edge processor, the Attention Chip, that captures and structures real-time gaze, blink, and pupil signals locally, eliminating cloud round-trips. The chip was demonstrated live at CES 2026, and the company is now developing an Attention Intelligence API and dataset products to help AI systems understand human attention at scale.
- Semiconductor
- Artificial Intelligence
- Developer Tools
- Hardware
Funding
Founders
Founder details are not available yet.
Product
Problem
AI systems currently lack real-time awareness of human attention and cognitive state, limiting their ability to respond naturally to users. Traditional approaches rely on cloud-based processing or lack structured eye-tracking data, adding latency and complexity while failing to capture meaningful behavioral signals needed for context-aware AI.
Solution
Attentix AI has built a dedicated edge processor, the Attention Chip, that captures and structures human eye data in real time. The chip converts raw gaze, blink, and pupil signals into clean, machine-readable outputs at the edge, eliminating the need for cloud round-trips. The company demonstrated this chip live at CES 2026, with the core signal engine fully operational. On top of this hardware foundation, Attentix AI is developing an API layer, structured datasets, and deployment infrastructure to enable companies to integrate human-state intelligence into their AI products at scale.
Target Audience
AI system developers, robotics companies, and product teams in automotive, consumer electronics, and human-computer interaction markets that require real-time human attention data at the edge.
Features
- Dedicated edge processor (Attention Chip) for real-time capture of gaze, blink, and pupil signals with low-latency local processing
- Core signal engine that converts raw eye data into structured, machine-readable outputs in real time
- Attention Intelligence API and dataset products in development to support software-layer integration
- Hardware-ready deployment model suitable for embedded and edge computing environments
- Demonstrated live at CES 2026, indicating technical maturity and field-readiness