English
Understand real speech, not just test questions
Words, intensive and extensive listening and pronunciation rules in one system. Every sentence and every miss is kept, and tomorrow's practice is picked from your real problems.
8,000 words
COCA's most frequent, reviewed as you forget
Academic listening → C2
An intensive path, one sentence at a time
Shows · podcasts
Real material, picked by how many words are new
American sounds
Linking and reductions, from what you miss
The learning loop, data-driven
- 1 Practice Dictation · shadowing · podcasts · word memory
- 2 Signals Each miss traced by what was written; the AI's, confirmed by you
- 3 Induce A family of rules learnt as one principle
- 4 Decide 4 missed most + 1 worst by rate; mended ones rechecked at 7/21/60 days
- 5 Picked Every track does them first
Against the usual way
At 2 hours a day, from B1 now, how long it should take · simulation calibrated on real data and research (central estimate) · usual = the old way on one day in three; V2.0.0 = every day by the plan: short-chunk dictation first, then dictation and podcasts
Usualabout 2.5 years · 3/8/2029
V2.0.0about 3 months · 1/25/2027×9.0 as efficient
Usual
Here
Textbook order
What
Re-ordered from your misses
Tests, impressions
Feedback
Every sentence, week on week
Forgetting unseen
Words
Reviews timed to forgetting
Textbook dialogues
Material
Real shows, podcasts