IB DP Math AI HL Tutor in Abu Dhabi

In the first four weeks an AI HL tutor should produce three things: a diagnosis naming the specific topics costing marks, a marked set of past-paper questions showing where the loss happens inside each answer, and a realistic read on the exploration. Nothing in that list is an opinion. All of it is a document, and all of it should be in a parent's hands before a second month is paid for.

The diagnosis is built from the student's own work rather than from a bought diagnostic test. A tutor asks for the last two school assessments, the most recent mock if there is one, and whatever exploration material exists. Those scripts are then re-marked against the IB's published command terms, because the gap between what a student knows and what a student scores is usually a command-term gap. A question that says 'find' wants a value. A question that says 'hence find' wants a value derived from the previous part. A question that says 'comment on' wants a sentence about the context, and a correct number with no sentence scores nothing on that mark.

The marked set then gets sorted into two piles, and the sort is the whole point. One pile is content the student has not learned: transition matrices, say, or Euler's method. That pile is fixable by teaching and the timeline is predictable. The other pile is content the student knows and still loses marks on: rounding before the final step, a calculator left in radians, a chi-squared conclusion stated without reference to the significance level. That pile is fixable faster and is worth more marks per hour of work, which is why it gets attacked first.

By week four a family should be able to answer a concrete question: which three topics are costing the most marks, and what is the plan for each. If a tutor cannot answer that from written evidence after four sessions, the match is wrong and should be changed rather than persisted with.

AI HL is assessed by three written papers in the examination session plus one internally assessed mathematical exploration. The three papers are not three versions of the same test. Paper 1 is a set of shorter questions covering the whole syllabus, Paper 2 is a set of extended-response questions, and Paper 3 is two long problem-solving questions that only Higher Level candidates sit. The exploration is marked by the school against the IB's five criteria and then moderated externally.

Technology is assumed in every AI paper, which is a structural difference from Analysis and Approaches. An AA HL candidate sits Paper 1 without a calculator; an AI HL candidate is expected to have a graphic display calculator on the desk for all three papers. That changes what the examination actually tests. Computation by hand is largely gone, replaced by model selection, interpretation and the ability to explain what a number from the calculator means in the context of the question.

The practical consequence is that calculator workflow is examinable technique rather than a convenience. Knowing which regression model to fit, how to run a hypothesis test and read the p-value, how to find an intersection numerically, and how to store and recall exact values instead of retyping rounded ones are each worth marks across all three papers. A student who is fluent with the syllabus and clumsy with the device loses marks that have nothing to do with mathematics, and those are among the cheapest marks to recover in a tutoring programme.

The table below sets out the components and the dominant failure mode in each. Weightings are the IB's published course structure for Applications and Interpretation at Higher Level.

Paper 3 exists only at Higher Level and carries a fifth of the grade in one hour. It presents two extended problem-solving questions, each heavily scaffolded: a context is set, a model is introduced, and the question walks the candidate through it in parts that build on one another. The mathematics inside any individual part is rarely harder than Paper 2 mathematics. The difficulty is structural, and that is exactly why it responds to practice.

The characteristic Paper 3 failure is abandonment. A student meets part (c), cannot see it, and stops — losing not only part (c) but parts (d) through (g), which depended on carrying a result forward rather than on having derived it. Examiners award marks for using a previous answer even when that answer was wrong, and students who have never been told this leave whole pages blank. Teaching a candidate to write down an assumed value, label it clearly, and continue is worth more on this paper than any additional content.

The second Paper 3 skill is reading the modelling instruction literally. These questions frequently ask a candidate to state an assumption, justify a choice of model, or comment on a limitation. Those marks are awarded for sentences, not for algebra, and students trained on Paper 1 habits write the algebra and skip the sentence. A tutor working through three or four past Paper 3 questions with a stopwatch usually finds the same two or three sentence-marks missing every time.

Because Paper 3 is HL-only, it is also the component a small school cohort is least likely to have drilled. An Abu Dhabi student in a six-strong AI HL class may have sat one timed Paper 3 before the mock. Four timed attempts with full marking is a realistic target for a tutoring block, and it is usually the single highest-yield intervention available in DP2.

The exploration is marked out of twenty across five criteria, and the mark distribution surprises most families. Presentation and mathematical communication are worth four marks each; personal engagement and reflection are worth three each; the mathematics itself is worth six. A student who produces impressive mathematics in a badly structured document with no reflection can score eleven out of twenty, which at HL is a meaningful dent in a final grade.

Criterion E, use of mathematics, is where Higher Level candidates most often lose marks, and the reason is level rather than accuracy. The rubric asks for mathematics commensurate with the course. An AI HL exploration built on a linear regression and a correlation coefficient is accurate and insufficient. One built on a transition matrix with a steady-state interpretation, or a numerically solved differential equation with a sensitivity check, meets the level the criterion describes. Choosing the right question at the start is therefore worth more than polishing at the end.

Criterion D, reflection, is the cheapest criterion to recover and the most frequently neglected. Reflection means commenting on the significance of results, on the limitations of the model chosen, and on what a different approach would have shown — distributed through the work rather than parked in a final paragraph. Two or three substantive reflective comments at the points where a decision was made will usually move this criterion.

What a tutor may do here is bounded by IB academic honesty rules, and the boundary is not blurry. A tutor may discuss topic viability, point at a criterion and say which descriptor the current version meets, flag a notation convention, and ask the questions that make a student notice a gap. A tutor may not supply text, supply the mathematics, or hand back a rewritten version. The work must be the student's from the first question to the last page, and the school will ask the student to confirm exactly that.

Mark loss in AI HL is patterned rather than random, and the pattern repeats across cohorts. Statistical inference accounts for the largest single block. Candidates run a chi-squared test or a t-test on the calculator, read off a p-value, and then write a conclusion that omits the significance level, omits the context, or states that the null hypothesis is 'true' rather than not rejected. The arithmetic is done by the device; the marks are in the sentence.

Modelling questions produce the second cluster. A question gives data and asks which model is appropriate, and the candidate fits the model the lesson was about rather than the model the data suggests. Exponential, logistic, sinusoidal and piecewise behaviours each have visible signatures in a scatter plot, and recognising them is a trainable skill that takes a handful of sessions. The same cluster contains unjustified extrapolation beyond the data range, which examiners mark as a limitation to be commented on, not ignored.

The third cluster is technical hygiene, and it is the one that most annoys families when they see it: a calculator left in radian mode for a degree question, a value rounded to three significant figures and then reused in a later part, a final answer given with no units, a graph drawn with no axis labels. These are not mathematical failures. They are habits, and habits change in weeks when someone marks them strictly and consistently rather than noting them in passing.

The fourth cluster is the HL-only content that small cohorts rush at the end of DP1: matrices and eigenvalues, graph theory algorithms, complex numbers used in modelling, and numerical solutions to differential equations by Euler's method. Students often meet these topics once, under time pressure, in the same term as the exploration. A tutor can re-teach them properly in the autumn of DP2, which is late but not too late.

A family paying for weekly tutoring should be able to see progress without asking for a verdict. The mechanism used here is a running file, and it has four parts: a mark log by paper and topic, a set of marked scripts with the lost marks annotated, a dated version history of the exploration, and a short written note after each session naming what was worked on and what the next session will fix. None of that is elaborate. All of it is the difference between a judgement and a record.

In weeks one to four the file fills with diagnosis. Old scripts come in, get re-marked against command terms, and produce a topic list ranked by marks lost rather than by syllabus order. The mark log starts with two or three honest baseline figures from the student's own school assessments, not from a flattering entry test. A parent reading the file at the end of week four should see which topics are being attacked and why those and not others.

From roughly week five to the end of DP1 the file should show movement in the mark log on the topics that were attacked, and a first exploration question that survives scrutiny. Movement will be uneven: technical-hygiene marks come back quickly, new HL content takes longer, and statistical interpretation improves in steps rather than smoothly. A flat log across eight weeks with no explanation is the signal to change something, and it is better to see that in writing in October than to discover it from a February mock.

In DP2 the file changes character. The mark log becomes a record of timed full papers rather than topic exercises, the exploration version history closes, and the session notes shift to the three topics still costing the most. By the final term a parent should be able to point at a graph of paper marks over time and at four marked Paper 3 attempts. That is what a claim of progress looks like when it is checkable, and it is what you should ask any tutor in Abu Dhabi to produce.

Published rates in the UAE tutoring market are scarce, so the figures below are drawn from providers that publish their own prices and from Bestimate's 2026 Dubai tutoring cost survey. They describe the market, not IB Gram's own fee, and they are quoted as the ranges those sources state. Rates in Abu Dhabi track Dubai closely, with in-home visits running at the upper end because the emirate's distances are longer.

Three things move the number for an AI HL booking. Level is the first: Higher Level commands more than Standard Level across every provider that publishes a split, because the tutor pool is smaller and Paper 3 competence is a narrower qualification. Format is the second: an in-home visit prices the tutor's travel into the hour, and in a city where a cross-town trip at 4.30pm can take forty minutes each way, that is a substantial share of what a family pays. Timing is the third: the final term before the May session is when demand peaks and availability is tightest.

The format comparison is where online delivery earns its place rather than being asserted. For the price of one in-home hour at the specialist rate, a family can often book two online hours with a tutor chosen for AI HL Paper 3 specifically rather than for living within a twenty-minute drive of Al Raha. That is the actual trade: a wider specialist pool and no travel premium, against the loss of a person physically in the room. For a subject where the examinable skill is interpreting output on a screen, the trade is a comfortable one.

IB Gram does not publish a price list on these pages, because the right number depends on level, frequency and how long is left before the session. A free 45-minute trial lesson with the proposed tutor comes first, with no card details taken, and the fee is confirmed in writing before any paid session is booked.

Private schooling in Abu Dhabi and Al Ain is overseen by the Department of Education and Knowledge, which u.ae describes as supervising the licensing and accreditation of private schools in the emirate. ADEK inspects schools and publishes ratings, and those ratings are worth reading before a school transfer. What ADEK does not do is license individual private tutors. That sits federally.

Private tutoring in the UAE is regulated by permit. The private tuition work permit was introduced by Ministerial Resolution No. 710 of 2023 and is administered by the Ministry of Human Resources and Emiratisation. According to u.ae the permit is granted free of charge for two years, and eligible applicants include teachers registered in government or private schools, employees in the government and private sectors, unemployed individuals, school students aged fifteen to eighteen, and university students. MoHRE's own service page, 'Issuance of a New Work Permit - Private Tutor Licensing', lists a two-working-day completion time and is the route to apply.

Two consequences matter to an Abu Dhabi parent. First, individuals offering private tutoring without a permit are, in u.ae's words, subject to fines and penalties under the applicable laws; no fine schedule is published, so no amount should be quoted to you by anyone. Second, Abu Dhabi has an additional restriction: press reporting on ADEK's professional-ethics rules indicates that private and charter school staff may not provide paid private tutoring to students enrolled at their own school, and that the MoHRE permit's code of conduct carries a similar bar. A family should verify the current wording against ADEK's own policy document before relying on it.

The practical reading: a tutor's own school teacher is the one person who usually cannot be hired, and a tutor who cannot discuss permit status is a tutor to ask more questions of. IB Gram holds no ADEK, MoHRE, IB or Cambridge accreditation and claims none; it matches students to independent tutors and states the delivery model plainly, which is online one-to-one video tuition for every UAE family.

The AI versus AA question reaches a tutor in two different forms. In DP1 it is a live decision, and in DP2 it is a worry about whether the decision already made has closed a door. Both deserve a direct answer rather than reassurance, because the stakes are real and the information is published by universities rather than guessed at.

Engineering, mathematics, physics and most quantitative economics degrees at competitive universities state Analysis and Approaches, frequently at Higher Level, in their published entry requirements. Some will accept AI HL; many will not, and the ones that will not say so explicitly in their prospectus. Computer science varies by institution more than any other subject. Business, management, psychology, architecture, life sciences, medicine in most systems, and the social sciences broadly accept AI HL at Higher Level without comment.

The instruction that follows from this is dull and necessary: read the entry requirements for the specific course at the specific universities on the shortlist, on their own admissions pages, in the year the student will apply. That applies equally to UAE institutions and to overseas ones, and it applies before the option form is signed rather than after. A tutor can help a student read a requirement precisely, and should decline to summarise a requirement the university states for itself.

For a student already in AI HL and worried, the honest position is that AI HL is a full Higher Level mathematics course with 240 teaching hours and a demanding Paper 3, and that the overwhelming majority of degree programmes accept it. Where a specific target course requires AA, the options are a switch early in DP1, a different course choice, or a route via a foundation year. None of those is a tutoring decision, and nobody should sell tutoring as the fix for a prerequisite mismatch.

Diploma students in Abu Dhabi sit the May session, which means the real calendar pressure arrives much earlier than families expect. Working backwards from the examination window rather than forwards from today is the planning habit that separates a calm DP2 from a frantic one, and it is the one thing no competitor page in this market seems to publish.

The fixed internal deadlines come first. Schools set their own exploration submission dates, typically with a first draft in DP2 autumn and a final version in the winter term, and those dates land in the same weeks as mock examinations. A student who starts thinking about an exploration question in January has lost the part of the process that Criterion C rewards, because genuine personal engagement takes time that a deadline has already consumed. The useful target is a viable question agreed before the end of DP1.

Ramadan sits inside this calendar and shifts earlier each year on the Gregorian calendar. School hours shorten, energy patterns change, and families in Abu Dhabi routinely move tutoring to post-Iftar slots for the month. Online delivery makes that trivial to arrange; an in-home arrangement often cannot absorb it at all, because a tutor's evening becomes a sequence of long drives. Planning the Ramadan schedule in advance, rather than cancelling four weeks, protects the DP2 revision run.

For the examination dates themselves, go to the source. The IB publishes its session timetables and results dates on ibo.org, and a school's Diploma coordinator will confirm the exact windows and the EUR and remark deadlines for that cohort. Cambridge and Pearson publish their own IGCSE timetables separately — Cambridge notes that every centre is allocated to one of six administrative zones and that using the wrong zone's timetable is the candidate's risk. No date on this page should be treated as a substitute for those sources.