Dr. Sarah Jenkins.
Eighteen years as a Cambridge IELTS examiner. Three Cambridge English research papers on IELTS Speaking band-descriptor calibration. The person who shapes every single rubric Sprechify scores against.
Cambridge, then us.
Dr. Jenkins joined Sprechify in 2024 after 18 years at Cambridge English. She has personally scored over twenty thousand IELTS Speaking sessions and trained two hundred examiners.
Her academic work focuses on the gap between rubric scoring and learner-perceived feedback — the reason most IELTS prep apps miss the point. At Sprechify, she rewrote our scoring system to match what an examiner's pen actually does on the page, not what a checklist says.
Three questions, three honest answers.
Why leave Cambridge for an AI startup?
Because the bottleneck has never been the rubric — it's been the cost of an examiner's hour. A learner in Karachi cannot pay for 50 hours of human practice. They can pay for unlimited hours of a system that scores the same way I do. The math is not subtle.
What do most IELTS apps get wrong?
They optimise for engagement, not accuracy. A real examiner would not reward you for a streak. A real examiner would tell you that your Part 3 answers all collapse into the same hedging structure — and then make you do twenty of them differently.
What changes about Sprechify with you in the chair?
Every rubric got rewritten in plain English. The feedback now reads like a marginal note a real examiner would leave, not a generated paragraph. And we calibrate every model release against a held-out set of human-scored sessions before it ships.