Key points
- Phasis uses graph technology to measure culture and map relations between people, values, and subcultures for AI comprehension.
- The World API provides a maintained, evidence-backed state of a domain, allowing agents to orient themselves against reality.
- The primary endpoint, /orient, returns relevant frames including current state, material changes, supporting evidence, and uncertainties.
Phasis Inc. has developed a new platform designed to help artificial intelligence agents orient themselves within complex environments. According to SiliconANGLE, the founders turned to graph technology to measure culture, aiming to map relationships not just between individuals, but also between people and their underlying values, beliefs, and subcultures. This approach allows agents to comprehend the composition of culture and make decisions based on its rapidly changing nature.
By the numbers
- One primary endpoint: The /orient endpoint serves as the main interface for sending tasks and receiving contextual World frames.
- One core asset: The World model is described as the durable asset that separates companies keeping agents current from those piecing together data each time.
- One interview source: CEO Brandon Farley discussed the platform’s role in enterprise agent orientation during a interview with John Furrier on theCUBE.
What the figures mean
The Phasis World API is built specifically for production agents in enterprise settings. A "World" represents the maintained, evidence-backed state of a specific domain. The company builds and deploys these Worlds so that agents can be properly oriented before work begins and remain current as circumstances evolve. Brandon Farley, co-founder and CEO of Phasis, noted that while reasoning capabilities in AI are scaling rapidly, expertise and understanding of the real world are not advancing at the same speed. He stated that Phasis addresses this gap by providing a representation of reality for agents to orient against.
Farley explained that the World API is designed to integrate into existing systems, providing the primitives needed to build this reality layer. He argued that the compounding benefits of agents oriented against reality will distinguish companies that adopt this approach from those whose agents must piece together information anew for every task. That suggests organizations relying on static data feeds may face operational lag compared to those using dynamic, evidence-backed orientation frameworks.
What should you do?
Managed service providers should evaluate how their current AI deployments handle contextual changes and domain-specific expertise. It is important to verify whether agents are accessing real-time, verified state information or relying on outdated training data. Organizations should assess the risk of agents making decisions based on incomplete or unverified cultural and operational contexts. Monitoring the performance of agents against known reality layers can help identify gaps in orientation and decision-making accuracy.
What to do and how to stay safe: Phasis
- Audit current AI agent workflows to determine if they have access to real-time, evidence-backed domain states before executing tasks.
- Evaluate whether existing systems provide mechanisms for agents to understand material changes and uncertainties in their operational environment.
- Monitor agent decision-making processes to ensure they are oriented against verified reality layers rather than piecing together fragmented data.
- Stay informed about emerging orientation frameworks like Phasis World API that may offer durable assets for maintaining agent expertise.
Step-by-step guide: Shadow AI Explained: 8 Questions Answered for Secure Operations
General security guidance from the BitShield News newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
What is the primary function of the Phasis World API?
The World API provides production agents with a maintained, evidence-backed state of a domain, allowing them to orient themselves using current state, material changes, and supporting evidence.
How does Phasis measure culture for AI agents?
Phasis uses graph technology to map relations between people, their values, beliefs, and subcultures, creating a measurement of culture that agents can comprehend and use for decision-making.
What is the /orient endpoint used for?
The /orient endpoint allows users to send a task to Phasis, which then returns the relevant World frames an agent needs to start, including uncertainties and supporting evidence.
