Evidence · Methodology

We don't publish vanity claims. We publish operating evidence.

AfroAI frames its proof as a pilot evidence pipeline — what will be published, when, and exactly how each metric is measured. Pilot data is published after milestone verification, not before.

Live curriculum data

Courses captured in the AfroAI database right now.

Live

565

Courses captured

Live

8

Faculties / departments

Live

2,156

Chapters captured

Counts are read directly from the curriculum database at page load and cached for 5 minutes. Only publicly visible textbook documents are counted. Browse a live course →

Pilot evidence pipeline

What will be published, when, and how each metric is measured.

Pilot data is published in a structured evidence pack after each milestone window closes. Every metric below is currently in the tracking stage — labelled “Pilot in progress” until verified.

AfroAI · Pilot evidence pipeline

Faculty activation rate

Pilot in progress

What it measures

The share of enrolled students in a pilot faculty who ask at least one question per week after course capture is complete.

Methodology

Computed from authenticated sessions in pilot courses. Numerator: distinct student IDs with ≥1 question in the week. Denominator: total enrolled students in the activated course cohort as reported by the institution.

Publication

Published monthly after each milestone window closes.

Weekly question volume per course

Pilot in progress

What it measures

The number of questions submitted across each active course in the pilot cohort, broken down by week.

Methodology

Aggregated from the question log with course-context filter applied. Excludes system-generated test queries. Published as weekly totals per course — not per student to protect privacy.

Publication

Published monthly in the pilot evidence pack.

Topic weakness map (pre-exam)

Pilot in progress

What it measures

The topics within each course that generate the highest question volume in the 3 weeks before a scheduled exam — a proxy for where students feel least prepared.

Methodology

Questions are tagged to curriculum chapters by the retrieval layer. High-frequency chapters in the pre-exam window are ranked and published. No individual student data is included.

Publication

Published once per exam cycle per pilot course.

Paid conversion from activated cohorts

Pilot in progress

What it measures

The share of students in a pilot cohort who upgrade from the free tier to a paid plan within the same semester.

Methodology

Attribution: a student is counted as 'cohort-attributed' if their account was created within the pilot period and their first course context was in the pilot faculty. Conversion: a paid subscription transaction recorded within the same semester window.

Publication

Published at end-of-semester milestone, once cohort window closes.

Honesty policy

What we commit to, and what we will not do.

No metric is published before the milestone verification window closes.

Every figure in a pilot evidence pack includes a methodology note linking back to this page.

Pilot-in-progress status is displayed explicitly until evidence is verified.

No invented testimonials. Every student story requires signed consent.

No fake partner logos. A school's mark does not appear until a pilot is signed.

Naira-native pricing. No USD on the consumer-facing site.

For procurement and research

Request the pilot evidence pack.

Request the pilot evidence pack

If you are a dean, HOD, registrar, or procurement officer evaluating AfroAI, you can request the current pilot evidence pack by email. The pack includes methodology notes for every metric and a summary of the current pilot scope. Evidence is only published after milestone verification — the email version is clearly labelled with what is verified and what is still in progress.

Request evidence pack — [email protected]