Skip to content
Trace

An AI course a school can run.

Material teachers assign, real models students use without an account of their own, and the controls a school needs. Everything on this page is in the product today unless it says otherwise.
Talk to the Trace team See the six-week unit

AI literacy is federal education policy.

The executive order Advancing Artificial Intelligence Education for American Youth made K to 12 AI literacy a national priority, with a White House task force, teacher training, and a student challenge. The order sets direction. It does not give a teacher anything to assign on Monday. Trace is that part.

Read the order. Trace is independent and not affiliated with or endorsed by any government agency.

What is in the library

Written for high school. Assign a whole chapter or one piece of it. Students get the version you assigned, so a later edit never rewrites work already underway.

Readings

Short articles that explain one idea at a time, with a check question after each. A student always knows what they just learned.

Labs

Change one setting and watch the result move. An idea sticks once a student has seen it change under their own hands.

Quizzes and unit tests

Every answer is kept, so a teacher can see which questions a class missed instead of a bare percentage.

Agent projects

A build that runs. A student submits the exact version and the runs behind it, so the work shows how it was made.

Agent Builder

Blocks on a canvas, connected into an agent that runs on a real model. No syntax stands between a student and the idea.

AI tutor

Help inside a lesson, step by step. It will not hand over a finished answer, so it stays a tutor and not a shortcut.

Dev tools

A real model API for students who want to write code. Credits come with the account, so nobody needs a card or a provider key.

Resources

Guides and documentation for teachers and developers, kept in one public place.

Browse the library

Fit it to the time you have

A teacher assigns a whole chapter, a single unit, or one activity. These are the three ways schools run it.

A unit inside an existing class

A few weeks in a CS, science, or media literacy course. Students leave with an agent they built and can explain.

A semester elective

The full sequence, from what AI is through data, models, language models, and agents, with a unit test at each step.

A school program

Several classes on the same sequence, starting with the teachers who volunteer, so the second group learns from the first.

The six-week starter unit

The unit most teachers run first. Every week ends with something a student made or wrote.

  1. Week 1

    What an AI system is

    Take a familiar product apart: its goal, its data, and how it fails.

  2. Week 2

    Data and labels

    Turn the world into a table and find out what the table cannot support.

  3. Week 3

    Errors and confident mistakes

    Set a threshold, read a confusion matrix, and see why base rates matter.

  4. Week 4

    Inside a language model

    Follow a prompt from tokens to an answer in the lab.

  5. Week 5

    How agents work

    Give a model a tool, run the loop, and read the transcript it produced.

  6. Week 6

    Build and submit an agent

    An open build in the Agent Builder, handed in with its runs.

Who does what

The workload by role, as the product exists today.

Students

  • Join with a class code. No roster import is needed to start.
  • Read, run the lab, answer the quiz, and build the agent.
  • Never hold an account with a model provider or pay for one.
  • Choose what to share. Nothing is public by default.

Teachers

  • Assign a chapter, a unit, or one activity with a due date, or schedule it to open later.
  • Preview anything as a student before assigning it.
  • See who has started and finished, with a plain reason next to anyone who has stalled.
  • Open a submission, read the answers, and follow every step the agent took.
  • Mark work reviewed and write a note. Extend a due date for one student.

Administrators

  • Create the school workspace and accept the Data Processing Agreement.
  • Invite teachers by email and manage seats. Teachers cannot create schools.
  • Pay one bill for the school.
  • On enterprise plans, sign in through your own single sign-on and pay by invoice.

What a teacher sees

Trace shows the evidence it actually has and leaves the judgment to the teacher.

Progress, with reasons

A student who needs a look is flagged with why: nothing opened this week, repeated failed attempts, or work past its due date. A bare count is never shown.

The exact submission

A project submission is the version the student handed in and the runs behind it. A teacher can see how the work was made, not only that it was.

AI activity

A teacher can read a student's AI conversations for their class and see any moderation flags. Oversight belongs to the adult in the room.

No grades

Trace records what a student did. It does not turn activity into a score, because activity is not the same as understanding.

The teacher workflow in detail

Before a student's words reach a model

Every model call on Trace takes the same path: the Agent Builder, the tutor, the dev tools, and the safety check itself.

  1. 1

    Identifiers come out

    A student's name, email, and other direct identifiers are removed before anything reaches a model, along with anything shaped like a phone number or address.

  2. 2

    The request says do not keep this

    Every request carries a zero-retention instruction, so the provider is told not to store it or train on it. Trace does not train on it either.

  3. 3

    Both directions are screened

    The student's message and the model's reply are each checked for school. Blocked content is withheld, and the teacher sees the flag, never the text.

  4. 4

    Tools are the ones Trace ships

    A calculator, a fact lookup, files inside the student's own project, and a few more. No open URLs, and no student code runs on Trace's servers.

The subprocessor list names every vendor that touches data. The Privacy Policy and the Data Processing Agreement carry the exact wording.

What it costs

Students in a class never pay. A teacher can start free. A school pays per teacher on one bill, with a shared credit balance for each class that refills every month. Districts and networks get a custom agreement with single sign-on and invoicing.

See current prices

How a pilot works

Pilots are free. We ask for a named contact, the teachers who will run it, and honest feedback.

  1. 1

    Pick the first teachers

    Two or three who volunteer, and the classes they will run it in.

  2. 2

    Run the six-week unit

    Each teacher sets up a class, shares the join code, and assigns week one. We onboard teachers on a call and stay in a shared channel.

  3. 3

    Review the term together

    We look at what students opened, finished, and built, then agree on whether to expand and to what.

Not there yet

Things schools ask for that Trace does not have. Ask, and we will tell you when.

  • Roster sync with a student information system or LMS. Students join with a code and teachers invite by email.
  • Lessons for younger grades. The material is written for high school, and chapter plans exist for younger bands.
  • Slideshows and question banks for each chapter.
  • Single sign-on for students. Staff sign-on is on enterprise plans. Students use email or Google.

Common questions

What schools ask before a rollout.

No. The Agent Builder is blocks and connections. A few advanced lessons include short Python examples, and the dev tools are there for students who want to write code.

Start with one class and the six-week unit.

A teacher can set it up today for free. A school conversation takes half an hour.

Create an account Talk to the Trace team