Private Beta · UK Education Market

Intelligent exam preparation,
powered by AI.

We're building the next generation of revision tools. Our platform uses advanced machine learning to understand exam requirements and personalise learning pathways for each student.

Our Technology

Cortex combines state-of-the-art AI with deep understanding of UK examination requirements to deliver personalised learning at scale.

Vision-Language Processing

Our pipeline processes handwritten answers and diagrams using large vision-language models (72B parameters), enabling accurate understanding of student work beyond simple text OCR.

Runs on Modal Labs GPU infrastructure

Knowledge Tracing

Bayesian models track each student's understanding across topics over time, predicting knowledge gaps and optimising revision schedules based on proven learning science.

Based on cognitive learning research

Mark Scheme Analysis

Our semantic engine understands examiner intent by analysing official mark schemes, going beyond keyword matching to identify the specific criteria examiners look for.

Aligned with AQA, Edexcel, OCR specifications

Proprietary Data

We've built a structured database mapping 10+ years of past papers to syllabus requirements, connecting 50,000+ questions by skill dependency rather than just text similarity.

Continuously updated with new papers

Built for the UK education system

Cortex is specifically designed for GCSE and A-Level students, with deep integration into UK exam board requirements and marking standards.

Personalised daily revision schedules
Instant feedback on practice answers
Spaced repetition for long-term retention
Progress tracking across all subjects
Works with existing past paper resources

Market Opportunity

£2.5B+
UK private tutoring market annually
5M+
GCSE & A-Level students each year
1/100th
Cost vs traditional private tutoring

Infrastructure

Our serverless architecture is designed for scale and efficiency.

GPU Compute

A100/A10G GPUs via Modal Labs for running 72B parameter vision models with sub-second inference.

Vector Search

High-dimensional embeddings for semantic question retrieval and knowledge graph traversal.

Event-Driven

Serverless processing pipeline that scales automatically with student upload traffic.

Interested in learning more?

We're currently in private beta. Get in touch to discuss partnership opportunities or request technical documentation.