UX
Designed for people
Interfaces shaped around human perception, cognitive load, and physical input.
The web was built for humans.
Agents are the new users.
Agent Experience (AX) is the discipline of designing digital products so AI agents can find, navigate, and operate them reliably.
task.status
recoverable
recoverable
AX is to agents what UX is to humans
For decades, UX defined how products serve people. Then developers became first-class users too, and DX emerged.
Agents are the third wave. They browse, read, authenticate, fill forms, call APIs, and complete tasks on behalf of people.
AX improves the same product rather than adding a separate agent layer.
DX
Designed for builders
APIs, documentation, SDKs, and tools shaped for people who code.
AX
Designed for agents
Structured paths and explicit states that agents can interpret and operate reliably.
For ecommerce operators
Apply AX to your storefront.
AXD shows whether agents can discover, understand, trust, and act on a website. Optagen starts with this AX diagnostic, then gives ecommerce operators the option to continue to a separate, read-only Commerce scan for margin, pricing, inventory, returns, demand, and marketing. Nothing changes without operator approval.
Audit your store
Free AX Score first · No account required
The framework
Five dimensions. One score.
| Dimension | Definition | Evidence |
|---|---|---|
| Discoverability | Can agents find your brand in category queries? | Machine-readable discovery |
| Navigability | Can agents accurately describe what you do? | Clear semantic structure |
| Operability | Can agents complete tasks using your product? | Explicit actions and states |
| Recoverability | Do agents self-correct when wrong about you? | Actionable failure paths |
| Transparency | Does AI represent your limitations accurately? | Readable constraints |
Task sequence
→→
Find the path. Interpret the state. Complete the task. Recover when conditions change.
The problem
Most products have an agent experience and do not know it
Every product already has an agent experience. The question is whether it was designed or left to chance.
Find
Agents cannot discover the relevant product or route.
Understand
Agents misread what the product does or where its limits are.
Act
Agents reach an interface but cannot complete the intended task.
Structured answers
Agent Experience, clearly defined
- What is Agent Experience?
- Agent Experience, or AX, is the discipline of designing digital products so AI agents can find, understand, navigate, and operate them reliably.
- How is AX different from UX and DX?
- UX serves people and DX serves builders. AX serves software agents that research, interpret interfaces, call tools, and complete tasks on a person's behalf. The three disciplines improve the same product for different users.
- What does the AX Score measure?
- The AX Score is a 0-to-100 diagnostic across Discoverability, Navigability, Operability, Recoverability, and Transparency. It evaluates agent experience. It does not measure revenue, margin, or financial performance.
- Is AX only about AI crawlers and search visibility?
- No. Discovery and accurate citation matter, but AX also covers whether an agent can interpret states, complete actions, understand constraints, and recover from errors.
- Does AX optimize a site for one AI model?
- No. AX favors durable web qualities such as semantic structure, explicit meaning, stable navigation, clear actions, honest constraints, and useful recovery paths across AI systems.
Audit your AX
The AX Score rates any product across five dimensions in ten minutes. No tools required.