Skip to content

BPPE Career

School Administrator

All sections
Ask Otto AI
Help

What is this page?

AI settings

Role aware

This page proves per-school AI controls, feature switches, model policy rows, and activation gates before live AI calls are enabled.

Audience
Administrator
Current limits
It uses local invented AI settings proof data only. It does not write school settings, call providers, create credentials, enable Private AI, enforce budgets, run evals, or activate student-facing AI.

Related Otto Guide links

Review AI settingstaskAI model policy referencereferenceAI school settings referencereferenceSettings, users, and rolesconceptOpen reference page placeholder

Support request boundary

Use Sev2 if AI could turn on without gates, send restricted data, or call a live provider; Sev3 if feature switches, model policies, retention, or Private AI mode are unclear; and Sev4 for labels or proof-copy feedback. This panel does not submit a support case.

Sev1: Critical outage or data-risk emergency

Target: 1-hour response, 24/7 paging during pilot

Sev2: Major workflow blocked

Target: 4 business hours

Sev3: Standard product question or non-urgent issue

Target: 1 business day

Sev4: How-to, cosmetic, or future enhancement

Target: 2 business days

These preview controls are placeholders. Real AI answers, search, walkthroughs, and support submission come in later tickets.

Notifications5

Static proof preview

Real notification feeds, delivery, and preferences arrive in a later ticket.

Sam Director

Sam Director

sam.director@otto-fixture.test

Switch local proof role

StudentInstructorSchool Administrator

This changes only the local shell proof. Production role changes remain server-validated future work.

Search is a keyboard-focusable skeleton. No real index or permissioned search results are implemented in this ticket.

Settings · AI settings

AI settings

Review per-school AI feature controls, model policy rows, and activation gates before live AI calls are enabled.

Otto Fixture BPPE Career School · AI settings

School Administrator · Sam Director

AI settings · Ticket 108

AI can be reviewed per school before it is turned on

This proof shows the school AI settings row, feature controls, model policy table, and activation gates. Everything remains mock-only until later AI safety tickets are complete.

FoundationProvidersRead toolsGuardrailsUsageEvalsPrivacyRetrieval

Global AI

Off

Production AI stays off

Features

5

Feature controls with policy keys

Policies

1/5

Only mock/local policy is active

Live providers

0

Anthropic, OpenAI, Bedrock, and Private AI are off

AI

AI remains governed by school policy

School AI settings are intentionally explicit: feature switches, model policies, and safety gates are visible before any real provider is connected.

Ticket 108

Global AI and student AI remain off for production; only local mock proof behavior is visible.

Otto Deep and Private AI cannot activate until ZDR, eval, PII minimization, notice, and budget gates are complete.

Every feature maps to one model policy key before it can call a provider.

Restricted data routes are explicit on each model policy and remain core-only, pseudonymized-only, private-only, or blocked.

Consequential actions still require a human click and cannot be completed by AI.

School AI settings

Global AI

Mock mode only

Review

School AI is visible in Settings, but production model calls stay off.

Otto Core

Enabled for deterministic proof

Enabled

Core can answer from local proof data and deterministic mock behavior.

Otto Deep

Disabled

Disabled

Deep handoff requires ZDR terms, evals, budgets, and PII minimization.

Private AI mode

Route pending

Blocked

Private AI cannot activate until the approved processing route exists.

Deep confirmation

Required

Review

Staff-facing Deep handoff will require visible confirmation before escalation.

Student AI

Off

Disabled

Student-facing tutor and study features wait for the AI-use notice and opt-out path.

Staff AI

Mock proofs only

Review

Staff can review settings, not call live staff assistants or tools yet.

AI-use notice

Not published

Blocked

Students and applicants need a plain-language notice before student AI turns on.

Feature controls

Ask Otto for staff

Staff · ask_otto_staff_questions

Mock enabled

Local proof can show source-linked answers without live models.

Policy

otto-core-staff-qa

Tier

Otto Core

Deep allowed

No

Human review

Not required

Student notice

Not required

Message drafting

Staff and instructors · message_drafts

Draft

Drafts require human send and notice language before activation.

Policy

otto-core-message-draft

Tier

Otto Core

Deep allowed

No

Human review

Required

Student notice

Required

Study with Otto

Students · study_with_otto

Off

Tutor work begins later; student AI remains off in Ticket 108.

Policy

otto-deep-tutor

Tier

Otto Deep

Deep allowed

No

Human review

Not required

Student notice

Required

AI grading drafts

Instructors · ai_grading_drafts

Blocked

Blocked until prompt-injection wrappers, evidence checks, and eval harness exist.

Policy

otto-deep-grading

Tier

Otto Deep

Deep allowed

No

Human review

Required

Student notice

Required

Compliance co-pilot

Compliance staff · compliance_copilot

Blocked

Private AI processing route is pending and no compliance conclusion may be automated.

Policy

private-ai-compliance

Tier

Private AI

Deep allowed

No

Human review

Required

Student notice

Not required

Model policy table

Core staff Q&A

otto-core-staff-qa

active

Only active because it resolves through mock/local proof data.

Task

staff_question_answering

Tier

Otto Core

Restricted route

core_only

Primary

mock · otto-core-deterministic

Fallback

none · none

Cache

no_cache

Max input

6000 tokens

Max output

700 tokens

Temperature

0 bp

Batch

Off

File retention

Off

School override

Locked

Eval passingai-core-staff-qa-golden

Core message drafts

otto-core-message-draft

draft

Needs notice copy, approval flow, and send controls before activation.

Task

message_drafting

Tier

Otto Core

Restricted route

core_only

Primary

mock · otto-core-deterministic

Fallback

none · none

Cache

prompt_cache

Max input

4000 tokens

Max output

500 tokens

Temperature

2500 bp

Batch

Off

File retention

Off

School override

Allowed

Eval pendingai-message-draft-golden

Deep tutoring

otto-deep-tutor

blocked

Blocked until ZDR terms, pseudonymization, evals, and student notice exist.

Task

student_tutoring

Tier

Otto Deep

Restricted route

pseudonymized_only

Primary

anthropic-shape · provider-key-required

Fallback

openai-shape · provider-key-required

Cache

no_cache

Max input

12000 tokens

Max output

1200 tokens

Temperature

4000 bp

Batch

Off

File retention

Off

School override

Locked

Eval blockedai-tutor-golden

Deep grading drafts

otto-deep-grading

blocked

Blocked until submission wrappers, quoted-evidence checks, and instructor approval flow exist.

Task

rubric_grading_drafts

Tier

Otto Deep

Restricted route

pseudonymized_only

Primary

anthropic-shape · provider-key-required

Fallback

openai-shape · provider-key-required

Cache

no_cache

Max input

14000 tokens

Max output

1200 tokens

Temperature

1500 bp

Batch

Off

File retention

Off

School override

Locked

Eval blockedai-grading-draft-golden

Private AI compliance narratives

private-ai-compliance

blocked

Visible as a placeholder only. Private processing route and counsel review are later work.

Task

compliance_narrative_drafting

Tier

Private AI

Restricted route

private_ai_only

Primary

private-route-pending · private-ai-route-required

Fallback

none · none

Cache

no_cache

Max input

16000 tokens

Max output

1600 tokens

Temperature

2000 bp

Batch

Off

File retention

Off

School override

Locked

Eval blockedai-compliance-private-golden

Activation gates

Zero-data-retention terms

Otto security

blocked

Anthropic or OpenAI ZDR terms must be attached before live use.

Student/applicant AI-use notice

School administrator

blocked

Plain-language notice and human-review path are not published.

Evaluation suite

Otto AI owner

pending

Golden tests are named but Ticket 112 will implement the runner.

PII minimization

Otto privacy

blocked

Restricted-data block tests and pseudonymization arrive in Ticket 113.

Usage budget

School administrator

pending

Budget and rate limits are displayed but enforced in Ticket 111.

Proof boundaries

This page does not write school AI settings, model policies, prompt versions, AI runs, budget rows, or audit events.

No Anthropic, OpenAI, Bedrock, embedding, batch, file API, provider credential, or live model response is active.

Private AI mode is a configuration placeholder until vendor terms, deployment route, and expert review are complete.

The model policy table values are local proof rows and are not production vendor pricing, routing, or compliance advice.

Source references

AI-R-001

Per-school AI controls

Each school can turn AI on or off globally and per feature before production use.

AI-R-002

Private AI mode

Private AI requires zero-data-retention or approved private routing and disables retaining APIs.

AI-R-014

Logged AI runs

Every AI run later joins model policy, prompt version, cost, model, user, and approval outcome.

PHASE-13

Ticket 108

Adds model policy table review and per-school AI settings before tools, validation, budgets, evals, and PII minimization.