Services
What SoAITrust does
Six services, delivered by our team with the support of the platform.
SERVICE 01
Certification verification at the source
ISO 27001 first. Then 42001, 27701, 27017/27018, 22301 and 9001, checked on IAF CertSearch.
- Status, number, issuing body and query date
- Certificate scope compared with the AI system's
- New check at every reassessment
SERVICE 02
Environment security
26 controls in 6 domains, from access to incident response.
- ISO 27001, NIST CSF, OWASP LLM, MITRE ATLAS, LGPD
- Prompt injection, AI-BOM, red teaming, secrets, agents
- Score per domain and maturity level
SERVICE 03
AI governance
21 controls in 6 domains, from policy to monitoring.
- ISO 42001, NIST AI RMF, EU AI Act, ISO 23894
- Risk classification, algorithmic impact, bias, model cards
- Human oversight and contestability
SERVICE 04
Risk and criticality
14 AI risks scored by likelihood and impact.
- 5×5 matrix and bands from Low to Critical
- Criticality score from 0 to 100
- Every risk linked to the gaps that support it
SERVICE 05
Due diligence report and scorecard
Verdict, P1 to P3 findings and an action plan with deadlines.
- 1-page scorecard and 15-to-20-page report
- Highlights, benchmark and rating details
- Valid for 12 months or until a relevant change
SERVICE 06
Agent red teaming with scenarios
The agent completes the legitimate task and refuses the attack. Tested before launch, protected from the findings, retested in production.
- Scenarios built from documented incidents and your sector's patterns
- Objective checker per scenario; blanket refusal counts as failure
- Guardrails written for your agent and a retest in 30 days