sAInaptic raises £1M seed to bring AI grading to apprenticeship assessments
sAInaptic has raised £1M in a pre-money seed round to take its AI auto-marking into apprenticeship assessments, Tech Funding News reported on Tuesday. The round values the Cambridge startup at £4.5M and was led by Cornerstone VC and Ufi Ventures, with participation from Elm and Brighteye.
sAInaptic is a female-founded AI startup whose product does one job: grade written work. Now it has fresh capital and a new target — apprenticeship assessments, one of the most labor-intensive corners of vocational education.
Grading to within a percentage point
The headline claim is accuracy. Per the report, sAInaptic’s grading AI “boasts almost perfect accuracy in assigning grades, with a deviation of less than 1%” — meaning its marks land within a point of a human grader’s. For a product that lives or dies on trust, the figure is the pitch: nobody adopts an auto-marker that grades badly, and assessors don’t get second chances with learners’ results.
From GCSE science to apprenticeships
sAInaptic was founded by Cambridge graduates Dr Rajeshwari Iyer and Kavitha Ravindran, who spotted a gap in GCSE science revision: around 70% of the questions are open-ended, yet most revision tools only handle multiple choice. Their auto-marker uses text analysis and natural language processing to identify correct concepts in a student’s written answer and return instant, teacher-like feedback — not a keyword match, but an attempt to understand what the student meant. By August 2024 the company reported more than 14,000 registered users across 1,000 schools in 20 countries, with the tool in use in over 50 UK secondary schools.
Apprenticeship assessments are the harder next step. Unlike school exams, they judge practical, open-ended evidence against competency frameworks — and the assessors doing that judging are stretched. Marking that takes days today is exactly where an AI that grades “almost perfectly” would pay for itself, if the accuracy claim holds in this new domain.
The Cornerstone bet
The round was led by Cornerstone VC, a Black-led pre-seed and seed fund that invests £250k–£1m into diverse founders, alongside Ufi Ventures — a fund focused on vocational and workforce technology, a natural fit for the apprenticeship angle. Elm and Brighteye joined the round. For a £1M raise, the investor mix reads like a deliberate strategy bet on where assessment is headed: out of the classroom and into the workplace.
The caveat
One note of honesty: everything about this round comes from a single source. Tech Funding News’ exclusive sits behind a paywall past its opening lines, and the under-1% deviation figure is company-reported — it has not been independently verified in public. The question an assessor should ask before adopting: measured against whose grades, and on which assessments? Until sAInaptic publishes the methodology, the number is a claim, not evidence. It’s a claim worth watching, though: if a startup can grade apprenticeship work to within a point of a human, assessment bodies across the UK will want to see the proof.
Sources: Tech Funding News — “Exclusive: Female-founded AI startup raises £1M to grade exams. Here’s the pitch deck” (October 13, 2026); Devdiscourse — “AI Start-Up sAInaptic Revolutionizes Teacher Marking in UK Secondary Schools” (August 30, 2024); Sifted — “Meet Cornerstone VC” (2022).