TRANSPARENT BY DESIGN
A useful next step, with a reason.
SkillSpira uses explicit rules, reported evidence and practical constraints. No AI model decides your stage or score.
Two complementary measures
AI Readiness uses the AIRI v3.2 behaviour-anchored self-assessment across five pillars. Overall score is the mean of all 15 answers; the five pillars have unequal question counts. We retain those framework scores and separately flag project prerequisites. An overall level does not grant deployment approval.
The Planner estimates capability from your reported evidence. A selected stage is a claim, not proof. Mismatches are explained. Readiness levels are never converted directly into technical stages. A linked readiness priority replaces one non-safety activity with targeted practice. Related safety checks can be grouped to fit the time budget; every safety skill remains required before agent execution.
Role → goal → evidence → prerequisites → project
- Practitioners can succeed through better individual work without building applications or recruiting a pilot team.
- Builders focus on working prototypes, realistic tests, integrations and operating safeguards.
- Leaders sponsor, prioritise and govern initiatives without having to build the tools personally.
- Non-public or uncertain data requires classification and appropriate permission before real-data experiments. Use synthetic examples while resolving access.
- Bounded agent work requires credential protection, logs, cost controls, permissions and human intervention individually. Persistent operation also requires terminal, environment, version-control, recovery and reliability evidence.
- Weekly time changes milestone count and scope. Three, six or up to nine activities are planned across 90 days; extend the timeline if prerequisites need longer.
Evidence and progress
Completing an activity, adding an evidence note and reviewing a measured outcome are distinct actions. Evidence notes are participant-recorded summaries, not independent verification or certification. Compare the same baseline measure before and after and include human review effort.
Readiness retakes are compared cautiously. Changes in question edition, content version or organisational response context are flagged. Score movement alone does not demonstrate causation.
Resources and learning support
Resources are matched to practical steps. The wider reading list labels commercial interests. Select a course, workshop or coaching only when it addresses a named gap; paid support is optional.
Tool names are examples selected for the project’s activity and operating constraints, not an exclusive or independently ranked list. Compare data handling, accessibility, cost and fit with your organisation’s approved alternatives. Vendors benefit when their products are adopted; vendor courses and documentation describe their own ecosystems. Resource cards disclose known commercial interests. SkillSpira does not establish that a named product is the only suitable choice.
Explore the reading list →