'%20fill='%2300FF00'/%3e%3c/svg%3e)
For educators
A real evaluation from every student. Not the four who fill the form.
Totem talks to every student who completed the cohort, in their own words. Which sessions landed, which dragged, and which to cut next time, ranked across the class.
Silent betaFully free in beta
Fall Cohort Evaluation, Data Science
Cohort end-of-term eval. Totem ran this conversation across 22 students and synthesized what came back into the takeaway below.
Key findingsAI-synthesized from 22 students
The applied-stats unit was the session 16 of 22 cited as most useful, most students used it in a real project within a month. The intro programming weeks ran too slow for 12 of 22 who came in with prior experience. Four flagged the guest-lecture format as flat (passive, no q&a). Eight asked for more case studies before midterms, and nine want the capstone broken into earlier checkpoints.
- Applied stats most-used
- Intro weeks ran slow
- Guest lectures flat
- Wants more case studies
- Wants earlier capstone checkpoints
- 20%Positive
- 40%Friction
- 40%Focus
- Responses
- 0
- Fully completed
- 0/0
- Avg. completion
- 0%
- Avg. length
- 0m
From silent-beta calls
“Every cohort, the same four students fill the form. Totem talked to all forty. The session I was certain was working was the one most students said dragged. I would have kept teaching it.”
— a course lead at a graduate program
What educators ask first
“But students will give bland answers to an AI.”
The opposite, in the silent beta. Students who'd write "great class" on a form give Totem the specific session that dragged, the moment they checked out, and what they'd cut next year. Anonymity loosens what the form-with-your-name forecloses. It's not the medium that gates honesty; it's the consequence.
What educators start with
The student conversations the form will never get.
End-of-cohort course evaluation
End-of-cohort evaluation
A real eval from every student. Not the four who fill the form.
Alumni outreach
Alumni outreach
What graduates say once the polite period is over.
Concept testing
Concept testing
Reactions to a rough idea before you build.
Why educators pick Totem
- 01
The whole class, not the four who fill forms. Eval forms select for the engaged and the angry. Totem reaches the quiet middle, the cohort that votes with attendance, not surveys.
- 02
Sessions ranked by what landed. Which session students remembered, which they referenced after, which they wished was longer. Curriculum decisions grounded in evidence, not the loudest TA's opinion.
- 03
Alumni five years out, on the record. Not what students say at graduation. What they say once the polite period is over and they've watched their career happen.
What changes
What changes for educators.
- 01
Coverage instead of selection bias
Everyone who completed the cohort hears from. Not the 10% who fill end-of-term forms. Curriculum signal grounded in the whole class, including the quiet ones.
- 02
Session-by-session granularity
Which session landed, which dragged, which everyone-but-you-felt-was-redundant. Specific enough to act on, not 'overall the course was good.'
- 03
Alumni feedback that stays honest
Five years post-graduation, the polite-period is over. Real signal on what prepared them, what didn't, and the gap they'd close in hindsight.
Common questions
What educators ask before they try Totem.
- How can educators run course evaluations that reach every student?
- By moving the evaluation from an end-of-term form to an async AI-moderated conversation. Totem talks to every student who completed the cohort — the engaged ones, the angry ones, and the quiet middle — and ranks sessions by what landed across the class, not what the four who replied wrote.
- What's the best tool for student feedback?
- Most student feedback tools are surveys (course-eval forms, end-of-term Likert scales). Totem runs depth interviews instead: the same questions across students, real probes on hedged answers, verbatims preserved. The feedback isn't a grade; it's an evidence base.
- How can educators avoid selection bias in course evaluations?
- Self-selection in eval forms is the largest unmeasured bias in course feedback. The four students who fill the form vote; the thirty-six who don't are the silent curriculum. Totem reaches every student in their own time, with structured questions, so the signal isn't dominated by extremes.
- How is Totem different from end-of-term course evaluation forms?
- End-of-term forms collect closed-ended answers and one open text box. Totem runs full conversations: the question adapts to the student's response, the follow-up probes hedged sentences ("the third session was fine"), and the synthesis ranks patterns across the cohort — which session landed, which dragged, which to cut next time.
- Can Totem help with alumni research?
- Yes. The strongest alumni signal arrives 5+ years post-graduation, when the polite period is over. Totem can run the same structured conversation with alumni cohorts at intervals — 1 year out, 3 years out, 5 years out — and produce a longitudinal signal on what prepared graduates, what didn't, and what curriculum gap they'd close in hindsight.
- How does AI-moderated student feedback compare to focus groups?
- Focus groups select for students willing to attend a focus group. AI-moderated interviews reach everyone — async, on the student's schedule, in their words. Focus groups can pursue follow-up threads in real time; AI-moderated interviews trade some of that flexibility for full-cohort coverage. Different tool for different research questions.
Your turn
Describe the cohort. Hear from every student.
One prompt covers the cohort. Totem talks to every student async. You wake to the session-by-session signal, ranked, with the verbatims that prove each call.
Drop your promptFree during the betaNo cardYour prompt, your data