senecaAIby Vindica
Menu
ProductDevelopersPilotTrustCompanyInvestorsRequest a pilot한국어
Seneca Starter KitPlainTechnical

Stop re-testing every new learner.

A student opens your app for the first time, and your app knows nothing about them. So it runs a placement quiz, or guesses. Often another app has already watched that student work. Seneca brings that work to you, with the student's permission.

Below is one made-up example, start to finish, in four steps. About three minutes to read. Prefer clicking through? Try the interactive demo.

Show me the exampleI'm an engineer: technical version
Test appsaw the student work
sends the facts
Senecakeeps them, with limits
student says yes
Tutoring appstarts informed
The two apps never talk to each other.

One student, two apps.

The student
A high-school student getting ready for the SAT
The test app
A maths test they took at school last week
The tutoring app
An app they're opening for the first time today

All three are made up, and so is every number below.

  1. 1

    The test app watches the student work.

    The student answers 16 algebra questions in a timed test. Now the test app knows how each question went, including a few things nobody outside it should see.

    Test app · results

    • Equations with one unknown4 of 5
    • Equations with two unknowns2 of 4
    • Linear functions2 of 3
    • Systems of equations1 of 42 hints · 1 skipped
  2. 2

    Private details stay behind. Only the facts leave.

    A small piece of Seneca software runs inside the test app. Before anything leaves, it sorts what the test app knows. Facts go out. The student's name is swapped for a code. The questions, the answers and the name stay where they are.

    Leaves the test app

    • Right or wrong, per question
    • Time spent
    • Whether a hint was used
    • The topic
    • The student, as a code: lrn_2261…

    Stays in the test app

    • The student's name
    • Their school grade
    • The questions
    • The student's answers
  3. 3

    Seneca files it, and keeps three things apart.

    What happened, what the test app thinks about it, and what Seneca reads from it are stored separately, so nobody mistakes an opinion for a fact. Here is one topic:

    What happened

    4 questions on systems of equations: 1 right, 2 wrong, 1 skipped. Hints on 2.

    The test app's opinion

    "Needs review." Filed under the test app's name.

    Seneca's reading low confidence

    Not solid yet. Only 4 questions, on one day.

    Never tested

    GeometryData analysisReading and writing

    Listed as "not measured". Never counted as weak.
  4. 4

    The tutoring app asks, then decides for itself.

    The student connects the tutoring app and says yes to one purpose: planning the first session. The tutoring app gets a short summary. It never sees the questions or the answers, and it never talks to the test app.

    What happens next is the tutoring app's choice. Another tutoring app could read the same summary and start somewhere else. Seneca doesn't pick.

    The tutoring app won't assume

    • An SAT score
    • Anything about geometry
    • How the student compares with others

Writing works the same way, with one difference.

In a maths test, "4 of 5 right" is a fact. In writing, a score is somebody's opinion: a teacher's, or an AI grader's. So Seneca files every writing score under whoever gave it, and the essays themselves never leave the app that collected them.

Here, a made-up student wrote two essays in an essay app, with two drafts each. A teacher and an AI grader scored every draft.

Use of evidence · out of 4

RaterEssay 1
draft 1
draft 2		Essay 2
draft 1
draft 2
Teacher1223
AI grader2334

Seneca keeps both rows and never picks a winner.

The writing tutor gets

  • Second drafts scored higher on both essays, from both raters
  • Use of evidence is where to start
  • Every score, with who gave it

It never gets, and won't assume

  • The essays, the prompts, the teacher's comments
  • Whether an essay was written by AI
  • Which rater is right

What each app gets out of it.

If your app receives

  • A first session that starts from real work, before any placement quiz.
  • Every fact arrives with where it came from and how sure anyone can be, so you decide how much to lean on it.
  • Something you can charge for: a premium start where students who connect their history skip placement.

If your app sends

  • You already measure. Seneca makes that count outside your app.
  • Your questions, content and teaching stay yours. Only facts leave, and only when the student says yes.
  • The student can bring history from other apps back to you the same way.

The kit has three more examples: the essay app and writing tutor above, an AI tutoring session handing over to a human tutor, and a vocabulary app handing over to a speaking app. Engineers can open all four, with the real files, a tool to test their own made-up data and a download, in the technical version.

Want to see it with your own app?

  • 1

    Ask for an integration review

    Tell us what your app records and which apps you'd want to hear from. We'll tell you what fits.

    Ask for a review
  • 2

    Send us a made-up example

    Send a made-up export in your real format. We'll show you what would leave your app and what the next app would receive.

    Send an example
  • 3

    Talk about a paid pilot

    A limited pilot connects one kind of event from your app to Seneca under a written agreement. The real integration format comes with it, in a sandbox.

    Talk about a pilot

We count visits and tool use anonymously: no cookies, no identifiers, and never anything you type or paste. Browsers that send Global Privacy Control or Do Not Track are not counted.

Please send made-up data only. We don't need real student records to review or map an integration.