Journal

A domain-by-domain coverage audit of StudyPRN's MRCGP AKT question bank: what it tests well, how its cognitive level maps to the exam, and the blind spots — statistics, currency an...

A clinician-led audit of AKTPrep's AI tutor for the MRCGP AKT: where its explanations track UK guidance, where the statistics and organisational strands need checking, and how to v...

PassMedicine does not currently publish an MRCEM SBA bank, so this is the honest finding plus a first-pass, review and exit method you can run on whatever RCEM-blueprinted bank you...

AI answer engines cite the medical exam content that gives them four things — a direct answer, named entities, cited evidence and a visible update date. This is the framework for p...

AI now scores structured answers and simulated consultations across medical exams — but an automated score is a measurement instrument you have to calibrate before you trust it. Th...

Geeky Medics' SCA AI scores your consultation across the RCGP domains — but automated scoring needs calibration before you trust it. This workflow shows how to separate observable ...

A Quesmed UKMLA workflow that sequences its many components — read, test, simulate — without re-encoding the same facts, audits both the AKT and CPSA halves separately, and keeps a...

A Medibuddy UKMLA plan built around the one scarce resource — unseen questions. First-pass, review and exit rules that keep the adaptive engine from burning your only honest readin...

Geeky Medics' AI patients only help if your case set mirrors the exam — this workflow shows how to build a balanced UKMLA consultation rota across the content map, preserve unseen ...

A commit-first workflow for Pastest's AI Tutor on the MRCGP AKT: reusable prompts for a currency-dependent exam, a statistics stress-test, a hard verification habit for UK guidelin...

A concrete PassMedicine workflow for the MRCGP AKT that keeps the evidence-based-practice and organisational thirds in view — baseline, three-strand floors, an error taxonomy and a...

Geeky Medics' SCA AI patients rehearse the 12-minute consultations the MRCGP SCA tests and a Q-bank can't — this simulator audit maps what it genuinely trains across the three RCGP...