It simulates every student to anticipate who finishes school, who reaches college, and what changes if you act.
Each dot is a student in Chile. Scroll to follow one ↓
15 years old, technical high school. Repeated 9th grade and is now in 10th. Orus knows the path, never the name.
Move up, switch schools, repeat or leave the system. Orus weighs each path with what it learned from 16 years of trajectories.
Adding up those futures, Orus forecasts each region and the country with its uncertainty range. It publishes only the totals, never a student record.
Illustrative student. Chile's regions from north to south, sized by their approximate share of school enrollment.
Tested on years that already happened
Given one student who left school and one who stayed, Orus flags the one at higher risk 92 times in 100. Using municipality averages, 75.
Accuracy on high school graduates in each Colombian department, predicting years the model never saw.
Times the real value fell inside the range Orus had given, in the department-level tests.
Use cases
With real data from Chile and Colombia. The same engine already runs in Mexico and Brazil.
Enrollment forecast by grade and territory, with its range: a line for the most likely outcome and a fan for what could also happen. Here, Colombia's high school graduates.
Grade 11 graduates. By department, the range held 8 times in 10 in the tests; for the national total it is indicative.
The twin learns what risk is made of. It is never one factor: it is being behind, grades and attendance together.
Probability of leaving the school system the following year. Chile, 2022-2025 transitions; the first row summarizes three groups of 2,600 to 4,700 students.
Orus crosses projected enrollment with teacher exits by age. You see how many to replace and where, before you open hiring.
Teaching positions 2028 enrollment no longer supports
Chile, nationwide. The enrollment drop is absorbed without layoffs by hiring fewer than leave; in each territory the balance can flip: where enrollment grows, every exit must be replaced. It is a planning indicator at today's hours per student, not a hiring forecast.
Test a policy before you launch it. Orus carries it through the whole system, from school to college.
Approximate scale from scenarios calculated by Orus (−20%: +9,349 graduates in 2028; +17,045 in 2030). In Chile, a single policy targeted at technical high schools adds +140 students to higher education in 2028. A scenario measures consequences; it is not a forecast.
Who asks these questions
First-year demand, retention, and which schools and regions your students come from.
Seats, teachers and dropout by region or department.
Enrollment by campus, early warning, and class and teacher planning.
Not a chatbot
It searches and summarizes what someone already wrote. If nobody published the figure, it doesn't know it.
Trust
Every figure says whether it was tested and with what range. If a model doesn't beat last year's figure, we publish that figure and flag it. What we can't measure well yet, we don't publish.
Trained on 16 years of trajectories. Every model is documented in a card with its data, its tests and its limits.
Orus publishes figures by territory, each covering at least 10 people. No names or identifiers.
Student-level records live in a private network with read-only access. Every source is logged with its version and fingerprint.
Integrations
Soon, Orus will live inside Edtools, next to your institution's daily operations.
Works with Claude, ChatGPT and Copilot through MCP. Every answer comes with its citation.
The same figures for your systems and dashboards, over REST with an OpenAPI spec.
Connects to your student information system, your ERP, your learning platform and your management dashboards.
FAQ
Orus is Edtools' digital twin of education: a model that simulates the next school year of every student in a country and adds up those futures to forecast enrollment, dropout, college entry and the teachers needed by region or country, always with its uncertainty range.
Official, anonymized records from education ministries: student trajectories, enrollment, graduates and transition to higher education. Every figure states its source, the year of the data and the model that calculated it.
No. Orus knows trajectories, never names or identifiers. It publishes only territorial aggregates of at least 10 people, and student-level records live in a private network with read-only access.
The same simulation engine is validated in Chile, Colombia, Mexico and Brazil. Hong Kong is in preparation.
A chatbot searches and summarizes what someone already published; if nobody published the figure, it does not know it. Orus calculates what will happen with models tested against years that already happened, and every answer comes with its model, its sources and its test.
Today, through a demo for universities, ministries, education departments and school networks. Modules inside Edtools are coming soon, and the MCP connection for AI assistants and the API are in early access.
We'll show you Orus with the data for your country and your region.