ICAO English Pro

The course is 13 modules on four tracks. Track A is radiotelephony: readback discipline first, then numbers, callsigns and clearances under pressure, then non-routine and emergency traffic, then degraded audio and accented radio. Track B is plain English, which is the half that decides most ratings: vocabulary and the review deck, the structures that carry reasoning, plain-English listening, describing a picture, and interview and discussion practice. Then 3 modules on the shape of the common test formats, and a capstone.

Each module names which of the six rated criteria it trains and which level band it is pitched at, so working through it is not a guess about what you are improving. Modules whose content is not published yet carry a single lesson that cannot be completed, which is what keeps the completion certificate honest: 9 modules gate it, and an unfinished module cannot be ticked off by accident.

The order matters more than the volume. Radio phraseology is a closed vocabulary you can drill to reliability, and it is the part most candidates arrive already half good at. Plain English under an unexpected question is open ended, it is where a rating is usually lost, and it cannot be drilled the same way. So the course spends most of its length on the second and uses the first to build the confidence that makes the second survivable.

The listening bank a Pro learner can reach is 194 exercises across 7 accent groups, and every Pro exercise exists twice: once clean, and once put through the degradation a real radio does to a voice.

The second version is the point. Accent practice on a clean studio recording trains the wrong thing, because the difficulty in a real transmission is rarely the accent on its own. It is the accent arriving through a narrow band, clipped at both ends, under squelch, on top of somebody else. A candidate who can follow a speaker in a quiet room and loses them on the radio does not have a listening problem in general; they have that problem, and it is trainable by hearing the same content both ways and noticing which words fell off.

The exercises are graded easy, medium and hard, and there is a fourth tier above hard that is deliberately harder than any real test presents. It is a training ceiling and it is never mixed into a format practice set, because putting it into one would teach a candidate that the format is harder than it is.

Three kinds, because a proficiency interview asks for three different things. Interview prompts, for the part where you talk about your own operation and the trap is answering in one clause. 12 picture scenes with model answers, for the part where you describe something in front of you and the trap is diagnosing it instead of describing it. Read-aloud drills, for the sounds your first language substitutes.

Every model answer is there to be compared with rather than copied. A model answer used as a script produces a rehearsed paragraph, and a rehearsed paragraph is the single easiest thing for a real examiner to detect, because the moment they ask an unexpected follow up the language collapses back to its real level. What a model answer is for is showing you the shape: where a reason goes, where a comparison goes, how somebody who does not know the exact word gets round it without stopping.

That last one has a name in the scale and it is worth learning deliberately. Paraphrase is scored as a skill, not as a failure. A pilot who does not know the word windsock and says the cloth cone that shows the wind direction has demonstrated exactly what the vocabulary criterion is looking for.

The deck holds 124 cards and each one carries a prompt, the model answer, and a teaching note that names the trap where there is one. Cards are scheduled by spaced repetition on your account, so a card you keep getting wrong comes back sooner and one you have secured stops taking your time.

It schedules separately from the ATPL question bank's own spaced repetition, deliberately: mixing aviation English vocabulary into a theory revision queue would put two unrelated kinds of recall in one interval schedule and neither would be scheduled well.

The vocabulary it drills is the working kind rather than the impressive kind. Weather and aircraft technical words, emergency and medical words, ground operations, and the paraphrase patterns that get you through the word you do not have. A test is not passed by producing rare vocabulary. It is passed by never being stopped by a missing word.

The three common test shapes are covered as 3 modules. Each one is a walkthrough of how that assessment is built and how to work through it, and a listening set sampled to that shape from the easy, medium and hard tiers. The speaking half of each is the mock interview run in that format's shape, so it arrives with the examiner rather than before it. What these modules teach is structure and strategy, which is the part a candidate can actually prepare and the part nobody publishes in a usable form.

They teach the format and nothing else. They do not reproduce any provider's questions, images, timing or scoring, they are not endorsed by or connected to any provider, and they cannot predict what a real sitting will ask you. SkyStudy is not affiliated with, endorsed by or approved by any language test provider or licensing authority.

Each module opens with a short film, and the heavier ones carry a deep dive as well: 16 in all. Every video lesson also carries a written introduction that stands on its own, so the lesson still teaches something on a connection that will not stream and on a deployment where video is switched off.

The certificate records that you completed a self-study course, and it says so on its own face rather than in a footnote on some other page. The disclaimer is part of the artwork, so a screenshot of the certificate carries it: not an ICAO Language Proficiency Endorsement, does not award, certify or predict any ICAO level, not issued by or affiliated with ICAO, EASA, any civil aviation authority or any test provider.

It has a public verification page and an id, which is the only thing that makes a completion certificate worth anything to a third party who did not watch you do the work. What it verifies is the completion, and nothing about your English.

The free trainer's self-assessment gives you a reading of the six skills and then forgets it, because it has nowhere to put it. On an account it is re-scored on the server, never trusted from the browser, and stored with the study plan it produces, so the next visit starts from what you did rather than from the top of the page.

The plan is ordered weakest criterion first rather than by your average, which is the right way round for an assessment awarded on the lowest of the six. It orders the course; it never locks any of it, so nothing a diagnostic says about you can put a module out of reach.

Every reading is kept, so the second and third can be compared with the first. None of them is a prediction of a real result and none is presented as one: it is a record of your own answers, scored the same way twice.

The capstone is a mock proficiency interview with an AI examiner. Three parts, 15 exchanges, 17 minutes of speaking time on a clock the server keeps rather than the page: 5 minutes of questions about you and your operation, 5 minutes describing a scene and being asked about it, and 7 minutes of open discussion about a situation you were not expecting.

An account gets 25 of them in its lifetime, and every bound on a session is enforced on the server: a candidate who edits the page, replays a request or leaves a tab open overnight still cannot get a sixteenth exchange or a twenty-sixth session.

The planner draws the scene and the discussion scenario from the ones your last 8 sessions did not use, and reaches for an older one only when nothing fresh is left, because a part with no opener would be the worse failure. One of part one's questions is a follow up picked to fit the subject you just raised, out of a bank written in advance rather than the next line of a fixed script, because a candidate who has rehearsed five biographical answers can otherwise walk through the whole part without producing an unplanned sentence.

This part opens when SkyStudy leaves beta. It is the one piece that costs money to run every time somebody uses it, so it is not something the project can give away while it has no revenue. Everything else on this page is open now.