AI your district can actually verify.

Most classroom AI is a black box – nobody can see how it decides, and nobody can promise a student is safe inside it. Olamis is built the opposite way: every promise is something you can check.

The problem you’re solving

One teacher with thirty students can’t notice every child who quietly gets lost – no matter how good they are. Nearly half of America’s high-school seniors score below basic in math. Olamis gives each student the one-on-one attention no teacher can give thirty kids at once – while keeping the teacher fully in charge.

Four promises, in procurement terms

• Primero, el permiso – consent-driven data handling aligned to parental/school authorization; designed to meet FERPA and COPPA expectations.
• Una persona decide – human-in-the-loop by design. The system recommends; educators decide. No automated high-stakes decisions.
• Sin calificaciones de máquina – no automated scoring, ranking, or labeling of students. Feedback stays with the educator.
• Sin secretos – explainable by design; every recommendation returns a plain-English rationale, and actions are auditable.

Deploys where your data must live

Olamis runs across cloud, hybrid, or fully on-premises / edge infrastructure – so student data can stay inside the building. The program never stores a student’s real name; students appear anonymized.

Designed to meet

Olamis is engineered to meet the standards districts require: FERPA, COPPA, ISO 27001 information-security practices, and WCAG 2.1 AA accessibility. These are design targets we build and test against as we move toward formal validation – not yet third-party certifications.

How to start

• Intro call – we learn your environment and data-residency needs.
• Scoped pilot – a defined group, with permission and teacher review built in.
• Review – you check the results, the privacy posture, and the explanations.