Singapore's FatFish AI builds Visual Action Intelligence software, led by its CAMP One engine, for cross-brand robot perception and action.
Fatfish Technology Pte Ltd, operating under the brand FatFish AI, is a Singapore-based company developing what it calls Visual Action Intelligence, a technology approach aimed at giving robots the ability to see, interpret, and act within unstructured real-world environments regardless of the underlying hardware platform. Rather than building its own robot chassis, the company positions itself as a software and algorithm provider that sits above existing robot hardware, supplying a unified intelligence layer that can be deployed across robots from different manufacturers and brands. This brand-agnostic approach is intended to let integrators and end users add perception, prediction, and autonomous action capabilities to machines that were not originally designed with advanced embodied intelligence in mind.
The company's flagship offering, showcased at the 2026 World Robot Conference in Beijing, is CAMP One, described in company materials as a Visual Action Intelligence Engine and as an embodied intelligence solution. According to the company's own exhibitor materials, CAMP One is structured as a layered software stack comprising a perception layer, a spatial intelligence layer, a decision-making layer, an execution layer, and a cross-robot execution layer, together intended to let a machine interpret its surroundings, anticipate what is likely to happen next, and translate that understanding into physical task execution. The company emphasizes that this intelligence layer runs through edge AI inference, meaning the perception-to-action pipeline is intended to execute close to the robot itself rather than depending entirely on a remote cloud connection, which is relevant for latency-sensitive tasks in dynamic physical settings.
FatFish AI's go-to-market approach centers on cross-brand compatibility: instead of tying its software to a single robotic platform, the company frames its value proposition around layering the same visual-action intelligence stack onto multiple types of robots already in service or in production from different manufacturers. This middleware-style positioning places the company in the software and algorithm segment of the robotics supply chain, adjacent to hardware makers rather than competing directly with them, and aligns it with the broader industry movement toward vision-language-action models and general-purpose embodied AI that can generalize across tasks and platforms.
The company exhibited at the World Robot Conference 2026 held at the Beijing Etrong International Exhibition and Convention Center, presenting CAMP One to an audience of robot manufacturers, systems integrators, and industry buyers under booth B413. Its corporate materials identify it as a Singapore-incorporated private limited company (Pte Ltd), consistent with a base of operations in Singapore, a jurisdiction increasingly used by companies targeting both Southeast Asian and broader Asia-Pacific robotics markets. Publicly available company materials at this time do not disclose the company's founding year, headcount, funding history, or manufacturing certifications, and no independent financial or investment disclosures were located during this review. Within the wider robotics supply chain, Fatfish Technology's activity centers on perception and control software rather than on manufacturing physical robot bodies, distinguishing CAMP One as infrastructure that other robotics companies could integrate into their own products rather than as a standalone consumer- or enterprise-facing robot.
Primary type & automation activities this supplier delivers:
Layered Visual Action Intelligence engine (perception, spatial, decision, execution layers) for cross-brand robot task execution via edge AI.
Contact Fatfish Technology Pte Ltd
WEBSITE
https://www.fatfish.aiHEADQUARTERS
Singapore
Company Facts
Founded
-
Primary Role
Software/Algorithm
Company Size
-
Primary Region
Asia-Pacific
Annual Sales
-
Funding Stage
-
Funding Total
-