The AI drafts. A teacher decides what runs.
Students draft protocols with the coach; a teacher reviews rigor and safety and approves what becomes runnable. Nothing a student writes turns into a real, hands-on experiment without an educator between the AI draft and the bench.
Everything a science class needs — alongside your LMS.
You get the class tools you’d expect — a gradebook, quizzes, discussions, curriculum, a calendar, messaging, and attendance — plus the science layer an LMS can’t give you: an AI protocol coach a teacher stays in front of, and reproducible experiment reports. It works alongside your LMS: today you export final reports to submit there, and LTI, Canvas, roster sync, and SSO are planned. Every feature below is available in the current release.
AI protocol coach, with teacher oversight
Sixteen task-scoped coaching agents help students turn a question into variables, controls, and a step-by-step protocol. Every draft is teacher-reviewable, and every run is metered and recorded to the school.
Auto-graded quizzes
Build quizzes with multiple question types. Students get instant scoring and feedback, and results flow straight into the gradebook.
Threaded discussions
Per-class discussion boards with threaded replies, so students can make and defend claims with evidence — not just drop an assignment in a box.
Gradebook, rubrics & standards mastery
A weighted gradebook with drop-lowest categories, reusable rubrics for consistent scoring, and standards-aligned mastery tracking across a class.
Curriculum & modules
Organize a course into modules and units, sequence content and assignments, and give students a clear path through the term.
Calendar, messaging & attendance
Assignments, due dates, and class events in one calendar; direct teacher–student messaging inside the workspace; and attendance tracked per session — all in one school-controlled place.
Verifiable certificates
Issue certificates for completed work that students — and anyone they share them with — can confirm at a public verification link.
Guided runs & reproducible reports
Students carry out an approved protocol as a guided session, log evidence step by step, and export a clean, reproducible report for grading, for a portfolio, or to submit in your LMS.
FERPA-minded privacy
School-controlled workspaces with no public student profiles, no ad targeting, and no data sale — and AI-training, marketing, and public sharing all off by default.
The inquiry-science loop, in six steps.
It follows the way a real research course already runs. Each step hands off to the next — assign, draft, review, approve, run, report — so the teacher stays in the loop and the student always knows what’s next.
Assign
Set a research goal, constraints, and a reusable rubric.
Draft
A student turns the question into a step-by-step protocol with the coach.
Review
You read it before anything is run; safety is flagged.
Approve
Only an approved protocol becomes runnable.
Run
The student runs a guided session and logs evidence.
Report
Export a clean, reproducible report for grading, a portfolio, or your LMS.
The science layer your LMS was never built for.
A general-purpose LMS is great at collecting work and posting grades — keep using it for that. Olto sits alongside it and adds the part a science class actually turns on: designing an experiment, keeping it safe and rigorous, and being able to reproduce the result. Deeper LMS integrations are planned, not shipped — today the hand-off is a clean exported report.
- Assignment drop-boxes and the grade column of record
- Files, announcements, and school-wide messaging
- Enrollment, rosters, and single sign-on today
- The gradebook your final reports are submitted into
- An AI protocol coach that shapes variables, controls, and procedure — with the teacher approving what runs
- A safety gate: dangerous topics refused, higher-risk ones routed to teacher review
- Guided experiment runs with step-by-step evidence capture
- Reproducible reports and verifiable certificates, plus standards-mastery tracking
LTI 1.3, a Canvas-friendly workflow, roster sync (SIS / OneRoster), and SSO / SAML are on the roadmap, not shipped. See the Education Trust Center for the honest integration status.
Sixteen task-scoped agents, with a teacher in the loop.
The AI coach is a set of task-scoped agents — not an open chatbot. Each one is configured to encourage teacher oversight and to refuse dangerous content. Every student prompt runs through a safety check before it becomes a protocol, and where it lands decides what happens next.
The student drafts a protocol right away — the coach proposes, and the student and teacher decide.
Routed to teacher review first, with safer alternatives suggested instead of a hard stop.
Designed to refuse dangerous topics outright, with a plain explanation of why, rather than offering a workaround.
- Explosives, energetic materials & pyrotechnics
- Toxic, choking, or chemical-weapon-type gases
- Pathogen culturing, select agents & gain-of-function work
- Controlled substances, drug synthesis & poisons
- Weapons, incendiaries & devices meant to harm
- High-voltage, high-energy & radioactive hazards
- Corrosive or violently reactive chemistry beyond a school lab
- Anything intended to injure a person, an animal, or oneself
This is an illustrative list of the categories the coach is instructed to decline or escalate — not an exhaustive catalog. Because screening is heuristic, a required educator review always sits between an AI draft and a real, hands-on experiment. Educators set the final boundary for their own classroom and jurisdiction.
The safety check uses patterns and policy rules, so it can produce false negatives — an unsafe request may slip past, and a benign one may be over-flagged. Treat it as a first filter, never as a guarantee.
A student draft is not runnable until a teacher reviews and approves it. Teachers can override the coach in either direction: reject something it allowed, or approve an appropriate activity it flagged.
Every hands-on experiment must be carried out under qualified adult supervision in an appropriate setting. The platform helps you plan and review; it does not supervise a lab bench.
Follow your school and district safety policies, your local regulations, and the required personal protective equipment and safety-data sheets for every material before any procedure begins.
Olto Discovery is a planning, review, and documentation tool — it is not a safety authority and does not replace professional judgment. A rigor score, a “safety reviewed” label, or a teacher’s approval never certifies that a hands-on procedure is safe to perform. Always follow your institution’s safety policies, applicable regulations, and qualified supervision — and stop any activity that appears unsafe.
Student data stays the school’s — and the risky defaults are off.
Olto Discovery for Education runs on school-controlled workspaces. There are no public student profiles, no ad targeting, and no sale of student data. The choices that usually put student data at risk are switched off unless a school turns them on.
Off by default. We do not train Olto-owned or public models on school or student data without the school’s explicit permission (prompts are processed by our AI inference provider under its API terms — see our education privacy commitments).
Off by default. Student data is not used for advertising or third-party marketing.
Off by default. A teacher must explicitly approve before any student work can be shared or made public.
Velora Biotech LLC aligns its education practices with FERPA, COPPA, California Student Privacy, and ATLIS guidance, and is pursuing an independent FERPA privacy assessment.
We do not claim certification under any of these frameworks. Read our education privacy commitments or the Education Trust Center.
One workspace, three seats at the bench.
The student does the science. The teacher keeps it rigorous and safe. The program lead keeps a whole cohort consistent. Everyone works from the same protocol, so nothing gets retyped between a lab notebook, a grading sheet, and a chat thread.
Students
- Draft a protocol with the AI coach and see variables, controls, safety, and materials
- Run it as a guided session and log evidence as they go
- Take quizzes, join discussions, and track their own progress
- Export a clean report, earn a verifiable certificate, keep a portfolio
Teachers & coaches
- Assign a goal with constraints and a reusable rubric
- Review every protocol before an experiment begins
- Grade with a weighted gradebook, run quizzes, take attendance
- Message students and give feedback without leaving the workspace
PIs & program leads
- Standardize how the group writes and documents protocols
- Build curriculum and modules a whole cohort moves through
- Keep student work in one school-controlled workspace
- Track standards mastery and a rigor score across the program
Request education pilot access
The platform is in active pilot, and we onboard schools in small cohorts so we can support each team directly. Tell us about your classroom or program and we’ll reach out as we open the next cohort.