AI-powered reasoning coach
GAMSAT preparation that trains
your reasoning, not your memory.
Verantrix continuously analyses how you think, identifies why mistakes happen, and coaches you to become a better problem solver.
The Problem
Students improve knowledge. They don't improve reasoning.
Most preparation optimises for the wrong thing, and plateaus for the same reason every time.
Doing thousands of questions
Volume without feedback on reasoning just repeats the same habits, right or wrong.
Watching videos
Passive content builds recognition, not the ability to reason under pressure.
Generic AI explanations
A general chatbot explains an answer once. It doesn't track whether you keep making the same mistake.
Repeating the same mistakes
Without a system tracking recurring patterns, the same reasoning error resurfaces attempt after attempt.
Why Verantrix Is Different
A closed loop from evidence to improvement
Every stage feeds the next, so your plan is always based on what actually happened, not a fixed syllabus.
- 1
Assessment
You attempt real questions under training or exam conditions.
- 2
Learning Evidence
Every answer, confidence rating, timing, and explanation is captured as evidence.
- 3
Diagnosis
Recurring reasoning patterns and error types are identified, not just topics.
- 4
Personalised Plan
Evidence becomes a specific plan for what to do next, and why.
- 5
Guided Mistake Review
Mistakes are corrected through a structured process, not just shown an answer.
- 6
AI Coaching
A tutor grounded in approved content helps you close the specific gap.
- 7
Follow-up
Corrected reasoning is checked again later, to confirm it actually transferred.
- 8
Adaptive Planning
Your plan adjusts continuously as new evidence comes in.
Two Learning Modes
Training and Simulated Exam never feel the same
One is built for learning. The other is built for realistic performance. Verantrix keeps them distinct on purpose.
Training
- Untimed and subject-based
- Every answer is analysed, not just marked
- You explain WHY you selected an answer
- AI coaching helps you correct your reasoning
Simulated Exam
- Real exam conditions, no interruptions
- Server-authoritative timing
- After submission, Verantrix flags incorrect answers, slow answers, overconfidence, repeated reasoning patterns, and requested explanations
Section 2
Essay feedback specific enough to act on
Most feedback returns one impression of the whole piece. Verantrix assesses seven dimensions separately, because an essay with strong ideas and weak reasoning needs different work from one with the reverse.
Prompt Engagement
building your response directly from the ideas in the prompt
Ideas & Development
developing your ideas further rather than adding more of them
Reasoning Depth
arguing your claims through rather than asserting them
Structure & Organisation
planning the shape of your essay before you write
Coherence
connecting your points so the argument reads as one line of thought
Clarity & Fluency
writing sentences a reader understands first time
Language Precision
choosing words that say exactly what you mean
Practice is chosen for you, then checked
Every essay you write updates your Section 2 profile. The profile decides what to practise next and which task to practise it on, so you are not left guessing which of the seven to work on first.
Working on one dimension does not always mean writing another full essay. A targeted drill takes three to ten minutes and works on that dimension alone. Drills are how you practise. They are not how improvement is judged.
An improvement counts only once it appears in a full essay. Getting better at a drill is not the same as writing better under exam conditions, and Verantrix does not treat the two as interchangeable.
The full sitting
Task A and Task B in one 65 minute block, with a single timer and no enforced split between them, matching how the examination actually runs. Pacing across the two tasks is part of what Section 2 tests, so Verantrix does not remove that decision by imposing two separate clocks.
Feedback covers each essay and the sitting as a whole. It identifies which dimension is limiting the writing and why. It does not predict a Section 2 score, and no marking service can tell you what ACER would award.
Reasoning Matters
Correctness and reasoning quality are not the same thing
A correct answer does not always mean good reasoning. An incorrect answer does not always mean poor reasoning. Verantrix measures both, separately.
The result you want: a correct answer reached the right way.
A sound process with a wrong outcome, usually one small correction away.
A correct answer that hides a fragile process, likely to fail under different phrasing.
The clearest signal for where coaching should focus next.
Guided Mistake Review
Correcting a mistake is a process, not a page of text
Every guided review follows the same structured sequence, so corrections actually stick.
- 1
Recall
You return to the question and recall your original attempt.
- 2
Retry
You attempt it again, before seeing any explanation.
- 3
Identify the mistake
You identify what went wrong in your own reasoning, not just the answer.
- 4
Compare with approved reasoning
Your reasoning is compared against an approved reasoning pathway, clearly labelled as such.
- 5
Correct reasoning
You work through the corrected reasoning step, not just a corrected answer.
- 6
Transfer question
A different question tests whether the correction actually transfers.
- 7
Future review
The mistake is scheduled for a later follow-up to confirm it stuck.
Personalised Planning
Evidence becomes a plan, automatically
Your plan updates continuously from what you've actually done, not a fixed, one-size-fits-all syllabus.
Today's plan
A specific, achievable set of activities for today.
Mistake reviews
Due reviews are prioritised so corrections don't go stale.
Targeted practice
Practice is assigned to your specific coaching targets.
Assessments
Periodic assessments keep your diagnosis current.
Follow-up
Corrected reasoning is checked again later to confirm it transferred.
Adaptive scheduling
Your plan adjusts automatically as new evidence arrives.
AI Coach
A tutor, not a chatbot
The AI coach never invents official answers. It behaves like a personal tutor, grounded in what's actually approved.
- Approved answers
- Approved explanations
- Approved reasoning pathways
- Your own evidence
- Your learning profile
- Your diagnosis
- Your coaching targets
Progress
Not a fake score prediction
Verantrix shows what actually changed: reasoning improvement, recurring mistakes decreasing, review completion, plan adherence, and reasoning maturity. Never a single fabricated number.
What most apps show
- A single overall score
- Predicted exam result
- Percentile rank vs other users
What Verantrix shows
- Reasoning improvement over time
- Recurring mistakes decreasing
- Review completion and plan adherence
- Reasoning maturity, tracked separately from correctness
Trust
Built to be trusted with your preparation
Evidence based
Every diagnosis and plan traces back to specific attempts, not assumptions.
Deterministic scoring
Correctness scoring is deterministic and consistent, not an AI judgment call.
Approved content
Answers, explanations, and reasoning pathways are human-authored and reviewed.
Privacy
Your learning data is used to coach you, not sold or repurposed.
Human-reviewed framework
The coaching and diagnosis framework itself is designed and reviewed by humans, not left to a model to invent.
FAQ
Common questions
Is this another question bank?
No. Question banks measure whether you got the answer right. Verantrix measures how you reached it (your confidence, your explanation, and whether the same mistake keeps recurring) and uses that to plan what you work on next.
Does AI decide the correct answer?
No. Correct answers, explanations, and reasoning pathways are approved content, written and reviewed by humans. The AI coach is grounded in that approved content plus your own evidence. It never invents an official answer.
Can it replace my tutor?
It is built to complement structured study, not replace a tutor or course you already trust. Verantrix specifically targets the part most preparation misses: identifying and correcting recurring reasoning mistakes.
How is this different from ChatGPT?
A general chatbot has no record of your specific mistakes, no approved-answer grounding, and no way to verify whether a correction actually transferred to a new question. Verantrix is a structured coaching system built around evidence, not an open-ended conversation.
Can repeat sitters benefit?
Repeat sitters often know the content but keep losing marks to the same reasoning pattern. Verantrix is designed specifically to surface the recurring patterns that more content alone will not fix.