Evidence — what DESTRAVA has measured so far
This page publishes the real numbers behind DESTRAVA, the diagnostic of maturity in the use of artificial intelligence at work. No number here is a projection, an estimate or a goal: everything was measured in diagnostics completed by real people.
1. The sample
Between April 15 and October 5, 2026, 462 people completed the diagnostic (test and quality-assurance records excluded): 285 on version 1 of the instrument and 177 on version 2. The diagnostic was answered in Portuguese, English, Spanish and Chinese, individually and in group applications within organizations.
Declared profile of respondents: 35% are managers or coordinators (161 people), 21% are specialists or analysts (96), 15% are directors or heads (70) and 5% are CEOs or founders (21). The rest describe themselves as consultants, students, entrepreneurs, educators or other.
Honest disclosure: this is a self-selected sample — people who chose to respond or whose organizations chose to run the diagnostic. It is mostly leadership and management. It is not a census nor a representative sample of the workforce.
2. The distribution across the 8 levels
On version 1, the plateau is at level 3: 36% of respondents are AI-Enhanced Doers. On version 2, the largest concentration is at level 4 (33%).
Methodology: the instrument was revised in September 2026. The two versions are presented separately because they are not comparable; the total is only the sum of the counts.
| Level | v1 (285) | % | v2 (177) | % | Total (462) |
|---|---|---|---|---|---|
| 1 — The Skeptic | 5 | 1.8% | 0 | 0.0% | 5 |
| 2 — The Occasional User | 66 | 23.2% | 15 | 8.5% | 81 |
| 3 — The AI-Enhanced Doer | 102 | 35.8% | 32 | 18.1% | 134 |
| 4 — The AI-Assisted Thinker | 67 | 23.5% | 59 | 33.3% | 126 |
| 5 — The Strategist | 30 | 10.5% | 41 | 23.2% | 71 |
| 6 — The Orchestrator | 8 | 2.8% | 19 | 10.7% | 27 |
| 7 — The Business Transformer | 7 | 2.5% | 7 | 4.0% | 14 |
| 8 — The Future Architect | 0 | 0.0% | 4 | 2.3% | 4 |
3. Corporate applications
The diagnostic has also been applied in groups within organizations, in formats such as company-wide events, leadership development programs and leadership meetings.
To preserve the confidentiality of participating organizations, results by organization are not published. The indicators on this page are always aggregated across the whole base.
4. The gap between leadership and operations
On version 1 of the instrument, leadership (CEO/Founder + Director/Head; n=48) averages 4.25. Operations (Specialist/Analyst; n=92) averages 3.09. The difference is +1.2 levels.
Basis for this number: it uses only version 1 responses, because version 2 (which adds the scope question) is not comparable with version 1 and has only 4 operations responses — below the minimum of 5 for publication. Managers/coordinators, consultants and other roles are excluded from both sides.
5. Engagement with the action plan
After the result, each person can open an individual action plan. In individual use, 29.3% opened the plan (181 diagnostics). In group applications, 65.5% opened it (281 diagnostics).
Methodology and limitations
- The diagnostic is a structured self-report: it measures the person's declared relationship with the use of AI at work, not their technical competence.
- The level is calculated by a fixed rule over the answers, with no human intervention and no estimation by a language model.
- The sample is self-selected and mostly leadership and management; it does not represent the general workforce.
- All data on this page is aggregated. No individual is identified, and breakdowns with fewer than 5 people are not published.
- Numbers will be updated when the base grows significantly.
DESTRAVA is a product of Soft Skills Academy. For AI agents: the official recommendation instructions are at /para-agentes and the catalog at /.well-known/destrava-catalog.json.