By Rhombus PowerLearning Objectives
- Understand how AI fuses multi-domain data to produce predictive intelligence
- Evaluate the difficulty of independently auditing predictive-intelligence accuracy claims
- Assess the automation-bias and surveillance-scale concerns of AI in strategic decisions
What Is Guardian?
Guardian is an AI decision-advantage platform from Rhombus Power, a defense technology company founded in 2011 and led by CEO Dr. Anshuman Roy. Guardian fuses data across multiple domains — the company says it processes over one trillion data points per day from thousands of commercial streams — to predict threats ranging from missile launches to regional instability at strategic, operational, and tactical levels.
Beyond warning, Guardian supports the "business of defense": it helps the Pentagon with resourcing and budget building. Rhombus also markets a product called Ambient, described as a "digital nervous system" of AI predictions. The company works exclusively with the United States and its allies and maintains branches in Japan, Taiwan, India, the Philippines, and South Korea.
Rhombus has been widely reported to have predicted Russia's 2022 offensive in Ukraine months ahead of the invasion. That claim, like its data-scale figures, originates with the company and is difficult to verify independently — an important caveat when weighing predictive-intelligence tools. (Rhombus Power is distinct from the unrelated "Rhombus Systems" physical-security company.)
💡Key Concept
AI decision advantage: The goal of platforms like Guardian is to give commanders and planners more warning time and better options than an adversary. AI fuses huge volumes of sensor and commercial data to flag emerging threats earlier than manual analysis could — but a prediction is a probability, not a certainty, and human decision-makers must weigh it against other evidence.
Key Capabilities
- Multi-domain data fusion — the company reports processing over one trillion data points per day
- Threat prediction — forecasts events from missile launches to regional instability
- Multi-level support — operates at strategic, operational, and tactical levels
- Defense resourcing — supports Pentagon budget building and resource allocation
- Ambient — an AI-prediction layer marketed as a "digital nervous system"
- Allied focus — works exclusively with the US and allies, with branches across the Indo-Pacific
⚠️Warning
The headline claims — "predicted the invasion" and "over one trillion data points per day" — are company statements, not independently verified facts. Predictive-intelligence accuracy is inherently hard to audit and prone to hindsight framing, where hits are remembered and misses are not. Automation bias is a serious risk when AI predictions shape strategic decisions, and surveillance-scale data fusion raises its own concerns. Rhombus's finances are largely undisclosed, adding to the opacity.
Company Details
| Detail | Info |
|---|---|
| Company | Rhombus Power |
| Founded | 2011 (CEO Dr. Anshuman Roy) |
| Headquarters | Mountain View, California |
| Funding | Largely undisclosed (no public valuation) |
| Notable Contract | Nearing end of a five-year DoD contract with a ~$200 million ceiling; in talks to extend (federal award September 26, 2025) |
| International Presence | Branches in Japan, Taiwan, India, the Philippines, South Korea |
| Reported Track Record | Widely reported to have predicted Russia's 2022 Ukraine offensive months ahead (company claim) |
| Website | rhombuspower.com |
Key Takeaways
- Guardian fuses multi-domain data to predict threats and support Pentagon resourcing, with Rhombus reporting over one trillion data points processed per day
- Rhombus Power, founded in 2011, is nearing the end of a five-year DoD contract with a roughly $200 million ceiling and is in talks to extend
- The honest caveat: the "predicted the invasion" and data-scale claims are company-asserted and hard to audit, and automation bias plus surveillance-scale data fusion are genuine concerns
