Platform
A structure for observing capabilities without losing the context that gives each signal meaning.
PRYSMAP starts with an organizational question, defines what needs to be observed, and considers the source and context of responses before building an interpretation. The goal is not to automate decisions, but to create better conditions for making them.
Architecture
Assessment begins before anyone responds.
Before a capability can be observed, it must first be defined in the context of the work: where it matters, what contribution is expected, and which behaviors would make it observable. Without that architecture, the same term can carry different expectations and produce comparisons that only look equivalent.
The platform organizes a versioned catalog of tracks, roles, levels, skills, elements, and expectations. Lineage makes it possible to trace the relationship among the model used, the questions asked, and the results obtained, even as the architecture evolves.
Distinction
Track, role, and level are not synonyms
A track represents a family of work; a role describes a contribution; a level expresses differences in autonomy, complexity, scope, and impact. Keeping them separate prevents every job title from becoming a new catalog and allows an expectation to change without redefining the person.
A useful architecture serves a decision. If it expands without a specific question, it can become harder to govern than the problem it was meant to clarify.
When it is still unclear which organizational problem should shape the question, our solutions can help identify where to begin.
Measurement
Each measurement answers a different question.
The platform offers four assessment modes. Each answers a different question and works at a different level of depth.
Explore
Pulse
Provides a quick, provisional interpretation of a skill. It helps identify where deeper assessment is warranted, but it does not confirm a final level or replace a more precise assessment.
Contrast
Check
Assesses each element of a skill through closed-ended questions and produces a more precise interpretation. If the available information does not support a confident conclusion for an element, it does not force one.
Focus
Target
Validates a skill against an explicit expectation tied to a role, context, or responsibility. It determines whether the expected level is supported, partially supported, not supported, or whether the available information is still insufficient to reach a conclusion, without assessing the entire skill.
Understand
Interview
Explores reasoning and context through open-ended questions, examples, and AI-guided follow-ups. It helps examine how someone applies a skill and how well they can support their responses with concrete situations.
Principle
The most useful measurement is not the longest one
The most useful measurement gathers enough information to answer the question and makes the limits of the conclusion clear.
Responses and context
What was observed must remain connected to the conditions of observation.
A perception is neither objective truth nor an isolated opinion. It is an interpretation shaped by a relationship, a moment, and a degree of exposure. Its value depends on knowing what the source was able to observe, which expectation applied, and how well they can support their response.
The platform retains organizational and session context, along with the model behind each question. When someone does not have enough information or concrete examples to support their response, they can say so rather than being forced into an answer.
Limit
Being unable to support a response does not mean a capability is absent
It may mean the behavior could not be observed, the source did not have enough visibility, or the question needs to be reframed.
Interpretation
A difference can mean more than one thing.
The distance between a result and an expectation does not reveal its cause on its own. It may reflect a gap, ambiguous criteria, insufficient information, or an expectation that no longer describes the work.
When an expectation has been configured, the platform can make a conditional, element-level comparison using minimum, desired, and ideal bands. This positions the result against a reference; it does not turn it into an automated decision.
Alignment
The result is consistent with the configured expectation. It can support a clearer interpretation, not an absolute truth.
Difference
The result and expectation do not align. It is worth reviewing the basis for the responses, along with the criteria and context, before assigning a cause.
Uncertainty
The available evidence is not enough to support a conclusion. In that case, gathering more evidence—or choosing not to act yet—may be the best decision.
A capability interpretation complements outcomes, performance, experience, and other signals. In high-impact decisions, it should be used alongside other relevant sources.
Responsible use
Clarity needs visible limits.
Purpose, context, and permissions define who may access and use each interpretation. Traceability preserves the connection among the model, question, response, and result so that any conclusion can be understood and reviewed.
- A signal summarizes what was observed from the available information; it does not define a person’s value or identity.
- A skill is not equivalent to overall performance, potential, availability, or willingness to move.
- Lack of assessment does not prove lack of capability.
- A comparison does not automatically decide a promotion, hire, or exit.
- A result should be used in proportion to its declared purpose.
When an interpretation is sufficiently supported to guide action, the platform can connect it to actions defined by the methodology: address a relevant gap or identify what information still needs to be observed. None of these actions amounts to an automated decision about a person.
The next step
Start with a question that needs better evidence.
Create an account and frame your first specific question, with an answer that can be supported and limits of use that can be explained.
Create an accountOpens the PRYSMAP platform in a new tab