How It Works

Three layers. One governed environment.

Orchen is not a chatbot with a dashboard bolted on. It's three deliberate layers: the AI environment students learn inside, the insight layer that informs the adults around them, and an optional device layer for schools that want hard enforcement.

01 Layer 01 · The AI Environment

A tutor that teaches, configured by your school.


Students experience Orchen as a conversation. Underneath: a Socratic posture hard-coded into the system architecture, adaptive presentation that adjusts to how each student thinks, subject-specific teaching modules, and assignments, flashcards, and quizzes that grow out of real sessions.

Student · Tutoring session
A live Orchen tutoring session where the AI guides a student through the difference between stars and planets with questions and a diagram

The Socratic loop: question, attempt, build. Not answer delivery.

Student · Writing workspace
The Orchen writing workspace: a full essay editor with a contextual outline card guiding the student through their draft

Writing assignments open a full workspace — the student writes, Orchen frames and checks.

Pedagogy rules

Encoded behavior, not marketing language

The Circle Rule switches approach after two wrong answers on the same concept. Partial examples stop before the final step so the student finishes. When the student has it, Orchen recognizes it and stops. Each rule is operational, documented on the platform page.

School configuration

Your identity, your defaults

The tutor's identity, tone, mechanical-feedback policy, and crisis routing are set at the school level. Every student meets the same configured tutor in every class — consistency is the point.

02 Layer 02 · The Insight Layer

Every conversation becomes structured signal.


A nightly pipeline turns sessions into learning profiles, concept-level mastery tracking, struggle and strength tags, and weekly plain-English advisor narratives. Teachers see their class as a distribution of understanding. Advisors get the synthesis. Parents get school-configured digests. Nobody reads raw transcripts — source conversations are deleted within seven days.

Teacher · Student insight
The teacher insight view for one student: best teaching approach, learning style percentages, struggle areas and strength areas by subject

A teacher's read on one student — built from sessions, not surveys.

Parent · Weekly digest
The parent weekly digest with a one-sentence summary of the week, activity stats, learning observations, and a suggested question to ask at home

The parent view: a clear weekly picture, ending with one thing to do tonight.

Crisis detection

Concerning content reaches a counselor, not an inbox

Every conversation is scanned. Detected concerns are flagged with a 24-hour urgency indicator, routed to the school's designated counselor, and tracked through a full lifecycle — open, reviewing, resolved, or escalated. Every step is logged.

Privacy by structure

Insight without surveillance

Students keep a private workspace — goals, journal, preferences — that no staff member can access. Staff see derived synthesis, scoped to their role, with every access written to an immutable audit log. The full model is in the Trust Center.

03 Layer 03 · The Device Layer (Optional)

For schools that want enforcement, not just policy.


Orchen runs in the browser on whatever fleet your school already has — no IT project required. For schools that want network-level enforcement, an optional managed device layer is available.

Default · No IT project

Browser, on your existing fleet

Orchen runs in the browser on the devices you already manage. Most pilots start here — nothing to provision, nothing to image.

Optional · Hard enforcement

Managed device layer

Devices provisioned with MDM management and DNS filtering that point to Orchen as the school's AI environment of record.

It exists for schools where a written AI policy has already failed and the board wants the next answer to hold.
Underneath It All

The data lifecycle, end to end.


A conversation is created, analyzed nightly into derived insight, and deleted at the source within seven days — leaving profiles and narratives governed by your school's retention policy, with crisis flags preserved and every staff access logged.

1
Day 0
Created

A student opens a chat. The conversation is scoped to the student and visible only to them.

2
Nightly
Analyzed into insight

The pipeline derives learning profiles, concept mastery, and weekly narratives. Crisis flags route to the designated counselor.

3
Day 7
Source deleted

The source conversation is deleted. Only the derived synthesis remains — nobody reads raw transcripts.

4
Ongoing
Insight retained per policy

Profiles and narratives persist under your school's retention policy. Crisis flags are preserved; every staff access is logged.

Read the full data lifecycle →

See It Live

Thirty minutes against the real product.

Every screen on this page is from the live platform. A walkthrough shows you the rest — with your questions driving it.

Book a walkthrough →

Or go deeper: the full platform architecture.