Campus Learning
Structured A1 to C2 and IELTS pathways with prerequisites, live cohorts, and progress tracked per module.
A1-C2six levels
- CEFR ladder
- IELTS track
- Prerequisites
Live cohorts, AI grading, one mistake ledger
Courses, peer practice, AI grading and certified teachers, unified by one ledger that turns every mistake you make into your next lesson.
Chapter 2: the tense that links past to now
Present Perfect
tutor ready
An invisible tutor who knows your mistakes. It is already playing; pause any time to ask.
Structure, people, feedback, and a human backstop. Each one does the job the other three can't.
Structured A1 to C2 and IELTS pathways with prerequisites, live cohorts, and progress tracked per module.
A1-C2six levels
Roleplay, debate, and peer critique. Live, async, or with an AI partner when a seat is empty.
28activity types
Rubric-mapped feedback the moment you finish, with the evidence, the distractor analysis, and why each answer is right or wrong.
112tracked skills
When the model isn't confident, the work routes to a certified teacher instead of guessing at you.
<24hreview turnaround
Every surface tags your work in the same 112-skill vocabulary, so a mistake you make in speaking is what your flashcards drill tonight.
Noticed
Grading tags the slip as Inference and files it.
Taught
The blackboard rebuilds today's lesson around it.
Drilled
A flashcard is scheduled for the day before you would forget.
Cleared
Inference comes back in a mock exam, and your band moves.
A live room around a shared canvas. Vote on what to fix first, and the corrections file straight to your ledger.
Dot vote
Weakest paragraph?
Everyone drops a dot, and the group converges on what to fix first.
Prompt, locked
Some people think students should pay university fees. Discuss both views.
In nowadays, many student thinks that the university fees should be payed by goverment.
Room and review
In the room (4)
Filed to the ledger
“is many reason” → “are many reasons”
Drop a correction sticky on a peer's screen, then send it to the ledger. It files as a MistakeEvent and feeds tomorrow's remedial pack.
The ledger files every correction by micro-skill, then feeds it back through whichever surface you open next.
Your screen on the left, your partner's on the right, and every slip becoming tomorrow's lesson while you practise.
You · interviewer
AI · candidate
Every slip on either screen files as a MistakeEvent. The Remedial Composer assembles a lesson in the background, while you practise.
Live mistake feed
Nothing yet. Slips land here the moment either of you makes one.
composer idle · waiting for events
You interview, your partner answers with a STAR coach guiding them, then you swap.
Six cohorts run in parallel, one per CEFR level, an hour a day with a certified teacher. Start on any weekday.
Foundations
survival mechanics · phonetic modelling
Day 34 of 50
19 of 25 seats taken
Momentum
breaking speaking anxiety · roleplay
Day 12 of 50
23 of 25 seats taken
Fluency Unlock
spontaneous paragraphs · no translating-in-head
Day 47 of 75
25 of 25 seats taken
Fossil Breaker
eradicating fossilized errors · ledger debates
Day 88 of 100
24 of 25 seats taken
Command
register, pragmatics · executive workshops
Day 61 of 125
17 of 25 seats taken
Mastery
rhetoric, native parsing · sparse but hyper-dense
Day 8 of 150
6 of 25 seats taken
One certified teacher per cohort, one hour a day, five days a week. Seats are capped at 25 and a new section opens the moment one fills, so nobody waits for a term. Join mid-cycle and the days you missed roll into the next one, with the whole recordings library yours from the first day.
Numbered edits in the margin, a band for each criterion, and a micro-skill tag on every slip.
Your Task 2 essay, graded as you write. Numbered inline edits link to margin comments, each tagged to a micro-skill and a band impact, scored by our RAG on the Cambridge descriptors.
Task 2 · graded by our RAG on Cambridge descriptors
Task Response6
Position clear; 2nd idea underdeveloped.
Coherence7
Logical; linkers mostly accurate.
Lexical6
Adequate; word-choice slips.
Grammar6
Verb-form & article errors.
Every module reads from the same practice, so your radar and your predicted band move the moment you finish a task.
14 days
2,860 XP to Level 3 · unlocks site-50 leaderboard
+1.0 band in 9 weeks · target 7.0 in 3 wks
▲ writing coherence +0.3
▼ R-04 inference −0.4 · V-01 −0.3
“I have went” → have gone
G-07 · seen 4× · Writing report
“do strict laws” → enforce
V-01 · seen 2× · P2P debate
“tink” → think /θ/
P-06 · seen 5× · Speaking
P2P match: Ayşe
wants a live roleplay
Writing report ready
6.5 · 6 comments
23 cards due
3 from your mistakes
Nine funded rivals, each better than us at the one thing they were built for. None of them covers all six.
| Player | CEFR ladder A1 to C2 | IELTS preparation | AI grading all four skills | Live human teaching | Community practice | Shared mistake ledger |
|---|---|---|---|---|---|---|
| Wordrig2026 · launch | Full | Full | Full | Full | Full | Full |
| Duolingo2011 · ~$5B+ cap | Partial | Not offered | Partial | Not offered | Partial, their strongest axis | Not offered |
| Preply2012 · $1.2B | Partial | Partial | Partial | Full, their strongest axis | Not offered | Not offered |
| EF Education First1965 · ~$3B+ rev | Full, their strongest axis | Partial | Partial | Full | Partial | Not offered |
| Cambly2012 · ~$60M+ | Not offered | Not offered | Not offered | Full, their strongest axis | Not offered | Not offered |
| E2Language2013 · test-prep | Not offered | Full, their strongest axis | Partial | Full | Not offered | Not offered |
| Magoosh2009 · ~$2-5M | Not offered | Full, their strongest axis | Not offered | Not offered | Not offered | Not offered |
| Cathoven~2020 · CEFR engine | Full, their strongest axis | Partial | Partial | Not offered | Not offered | Not offered |
| Engnovate~2023 · bootstrapped | Not offered | Partial | Partial, their strongest axis | Not offered | Not offered | Not offered |
| SmallTalk2Me~2020 · seed | Not offered | Partial | Partial, their strongest axis | Not offered | Not offered | Not offered |
Stitch it together yourself
$110-190/mo
Five separate logins, five paywalls, and zero shared memory. A slip in one tool can never reach the others.
One campus, one login
from $47/mo
One Mistake Ledger feeding every room. That coherence is the thing a stitched-together stack cannot copy at any price.
No subscription. Set where you are and where you're going, and one pack covers the whole pathway.
Every tier runs the same grading engine. The higher ones add live humans, frequency and priority.
Try the whole campus for a week.
$0
7-day full trial
The solo AI campus.
$549.91 once
Learn alongside other people.
$733.24 once
Everything in Foundational, plus:
A live class every day, at your level.
$1,016.57 once
275 journey days plus the 30-day life buffer
Everything in Smart Money, plus:
The same grading engine on every tier. The paid ones add live humans, frequency and priority. Open any row for the detail.
| Feature | FreeTry the campus for a week | FoundationalThe solo AI campus | Smart MoneyLearn with other people | PowerhouseA live class every day |
|---|---|---|---|---|
| Curriculum & Library | ||||
| 6 intro submodules · +6 band-matched after placement | Scaled to your journey | Scaled to your journey | Scaled to your journey | |
| No runway: unlocked content only | Journey days, exact | Journey days, exact | Journey days + 30-day life buffer | |
| 2× your journey days | 2× your journey days | 2× journey days, absorbs the buffer | ||
| Your monthly rate · one-time | Your monthly rate · one-time | Your monthly rate · one-time | ||
| Your unlocked content, forever | Free, forever | Free, forever | Free, forever | |
| Your unlocked submodules only | Unlimited | Unlimited | Unlimited | |
| 5 branches, lifetime | Unlimited | Unlimited | Unlimited | |
| Bundled with your unlocked submodules · one pass | Unlimited | Unlimited | Unlimited | |
| Assessment & AI Grading | ||||
| 1 full mock/mo · whole or sectional · timed only | 1/week scheduled | 2/week scheduled | 2/week + surprise mocks | |
| 1 lifetime | 1/month | 1/month | 1/month | |
| 1 WT1 + 1 WT2 /mo | Unlimited (custom RAG) | Unlimited (custom RAG + SpeechAce) | Unlimited (custom RAG + SpeechAce, priority) | |
| Mock's 3-part section + unlocked speaking questions | Unlimited (Azure) | Unlimited (Azure + SpeechAce) | Unlimited (SpeechAce primary, Azure backup) | |
| Mocks + unlocked practice questions | Unlimited | Unlimited | Unlimited | |
| Azure: mocks + unlocked speaking questions | Azure full suite | Azure + SpeechAce | SpeechAce + IELTS-rubric mapping | |
| Quaternary | Tertiary | Secondary | Primary | |
| Personalized Learning | ||||
| 1 lifetime | 1/month | 2/month | 4/month (weekly) | |
| 1/month | 2/month | 2/month | ||
| Last 30 days viewable | Full history | Full history | Full history | |
| Top 3 only | Full | Full | Full | |
| Today's band + confidence range | Trajectory model | Trajectory + rubric | Full ensemble | |
| P2P & Practice | ||||
| 9 lifetime, one per template | Full library · fair use | Full library · fair use | Full library · fair use | |
| AI + Live · low matchmaking priority | AI + Live + Async | AI + Live + Async | AI + Live + Async · priority | |
| 9 of 58, one per template | All 58 | All 58 | All 58 | |
| Live Classes & Community | ||||
| 1 hr/day (Mon-Fri) | ||||
| View-only | Fully interactive | Fully interactive · priority Q&A | ||
| Site-wide · read-only | Site-wide · read + post | Site-wide · read + post | Site-wide · read + post | |
| Gallery only: watch + chat | Stage + gallery · band-filterable | Stage + gallery · band-filterable | ||
| Private stats only | Band-filtered · both XP tracks | Band-filtered · both XP tracks | ||
| Human Escalation | ||||
| 2/mo: 1 writing + 1 speaking | 4/mo: 2 writing + 2 speaking | 8/mo: 4 writing + 4 speaking | ||
| Accountability | ||||
| Re-engagement nudges only | AI check-ins | AI check-ins | AI check-ins | |
| Streak only | ||||
Eight people who sat the exam, in their own words.
I had been writing the same article mistakes for two years and nobody had ever told me they were the same mistake.
The speaking partner never once let me get away with the answer I had rehearsed.
I work nights. Being able to join a cohort class on a Tuesday and again on a Saturday is the only reason I finished.
The written feedback reads like a teacher marked it, because a teacher checked the parts the model was unsure about.
I paid once, in naira, at a price that made sense here. No card getting charged again every month.
My first course told me my writing was weak. This one told me which four things were weak and in what order to fix them.
Practising with another learner sounds worse than practising with a teacher. It was not.
I stopped translating in my head somewhere in the second month and I could not tell you exactly when.
I stopped translating in my head somewhere in the second month and I could not tell you exactly when.
Practising with another learner sounds worse than practising with a teacher. It was not.
My first course told me my writing was weak. This one told me which four things were weak and in what order to fix them.
I paid once, in naira, at a price that made sense here. No card getting charged again every month.
The written feedback reads like a teacher marked it, because a teacher checked the parts the model was unsure about.
I work nights. Being able to join a cohort class on a Tuesday and again on a Saturday is the only reason I finished.
The speaking partner never once let me get away with the answer I had rehearsed.
I had been writing the same article mistakes for two years and nobody had ever told me they were the same mistake.
Short answers. If something here is still unclear, the report at the end of your placement covers most of it in context.
How accurate is the AI grading?
Every writing and speaking submission carries a confidence score. Above 92% the grade comes back immediately. Below it, the work routes quietly to a certified teacher rather than being guessed at, and how many of those reviews you get each month is set by your tier. Reading and listening never go near a model at all: they are marked against the answer key, so there is nothing there to be confident or unconfident about.