Edexia

Expert Feedback,for VCE English.Trained by 10+ VCAA AssessorsPaste your essay here... No account needed.Use exampleText ResponseCreating TextsGet expert feedback~2 min · No account neededTrusted bySt Bernard's CollegeKolbe Catholic CollegeSaint Ignatius CollegeHomestead SSCMarist College BendigoSt Bernard's CollegeKolbe Catholic CollegeSaint Ignatius CollegeHomestead SSCMarist College BendigoBacked byFeatured inTeachers stay in control1Edit everythingRewrite, delete, or build on any AI-generated comment before it ever reaches a student.2Add your voiceRecord a personal voice note on any piece of feedback.3Release when readyIf you set it to teacher review mode, nothing goes to students until you review and hit release.Accuracy you can trust.0181.2% exact. 98.3% within one.Across 579 essays at St Bernard’s College, Edexia matched teacher judgment at examiner-level consistency. When it’s not exact, it’s one grade band away.02Built on VCAA Rubrics, Calibrated by VCAA AssessorsEvery criterion, grade descriptor, and study design requirement is pre-loaded. A team of VCE English assessors continuously trains and validates the system against real marking standards.Privacy you can trust.01Your data stays yoursAll training data is siloed to individual or institutional accounts and remains the intellectual property of the account holders. We don’t use your data to train our models.02Stored locally & de-identifiedAll data is de-identified and securely stored on locally based servers.SOC2 IIISO 27001ST4SDetails All Texts. All Frameworks.And so much moreAI Detection & ReplayWatch a full replay of a student's writing process. See pastes, tab-offs, and AI-likelihood scores.Cross-Submission ReportsAI-generated summary of each student's strengths, weaknesses, and next steps across all submissions.Prompt & Stimulus LibraryHundreds of prompts filtered by text, theme, command term, and difficulty. Or paste your own.Benchmarking & CalibrationTeachers blind-grade the same scripts, align in a calibration meeting, then get graphs and AI advice on marking tendencies.Blind GradingDe-identified scripts randomly assorted to teachers. Enter your own grade before revealing the AI's.ModerationCross-check scripts, view the spread of judgments, and generate moderation reports with visualisations.Handwriting TranscriptionScans handwritten responses and transcribes them for easier review — substantially better than most human reading.And so much moreAI Detection & ReplayWatch a full replay of a student's writing process. See pastes, tab-offs, and AI-likelihood scores.Cross-Submission ReportsAI-generated summary of each student's strengths, weaknesses, and next steps across all submissions.Prompt & Stimulus LibraryHundreds of prompts filtered by text, theme, command term, and difficulty. Or paste your own.Benchmarking & CalibrationTeachers blind-grade the same scripts, align in a calibration meeting, then get graphs and AI advice on marking tendencies.Blind GradingDe-identified scripts randomly assorted to teachers. Enter your own grade before revealing the AI's.ModerationCross-check scripts, view the spread of judgments, and generate moderation reports with visualisations.Handwriting TranscriptionScans handwritten responses and transcribes them for easier review — substantially better than most human reading.Students can write, get instant feedback, and rewrite in the space of an evening — that cycle of improvement just wasn’t possible before.— Chris Mason, Head of English at St Bernard’s CollegeReady to try it?Sign up and try it out. No contracts or obligations, just experiment.Sign up freeFrequently Asked QuestionsIs AI grading accurate enough for VCE English?Edexia is the only AI grading tool purpose-built for VCE English, calibrated by a team of ten VCAA assessors. In trials across 579 essays, it matched teacher grades exactly 81.2% of the time and was within ±1 mark 98.3% of the time. Unlike generic AI tools, Edexia understands the VCE rubric, every text on the study list, and improves through school-level moderation.How does it handle different VCE texts?Edexia's calibration is largely text-agnostic, so the core rubric interpretation applies across texts. For each text on the VCE list, we build a knowledge base of themes, authorial intent, and key quotes, validated by our assessor team. Whether your students are studying The Memory Police or Bad Dreams, Edexia understands the text.What about data privacy?We treat your data like an LMS does. Everything is siloed to your school's instance, stored on Australian servers, encrypted, and de-identified before any AI processing. We never train on your data. We hold SOC 2 Type II certification and ST4S accreditation.How long until it saves me time?From the first submission. Even before Edexia is fully calibrated to your standards, it works as an AI scribe, drafting detailed feedback that you then edit and personalise. Over time, as you moderate and calibrate, the drafts get increasingly accurate and require fewer edits. Most schools report significant time savings from week one.What does onboarding look like?We start with a department meeting to demonstrate the platform and answer questions. Then individual 5–10 minute check-ins with each teacher to ensure they're set up, followed by fortnightly touch-points throughout the term. No formal training required. The platform is designed to be as intuitive as moderating a colleague.Can it detect AI-generated essays?Yes. Edexia includes AI detection built specifically for student writing. Unlike generic detectors that flag legitimate student work, ours is tuned to minimise false positives. We'd rather miss some AI use than wrongly accuse a student. Combined with writing replay (seeing how the essay was written keystroke-by-keystroke), you get a complete picture of authenticity. --- Our Company.Our StoryAt Edexia, we've always been driven by a desire to make a positive impact on education. We see the education system as simultaneously the most important and most underserved societal system. Our journey began when we co-founded a previous EdTech company in the tutoring and resources domain. While this venture taught us a lot, we found it hard to make the meaningful impact we had envisioned. This led us to pivot and co-found Edexia, focusing entirely on creating lasting, positive change in education.AI is going to fundamentally reshape education. We believe that should mean more personalised learning for every student and more time for teachers to do what only humans can: mentor, inspire, and connect. We're here to make sure that's what it looks like.Our BackgroundFounded in Australia by second-time EdTech founders, Edexia was incubated at Y Combinator, the Silicon Valley accelerator behind Airbnb, DoorDash, Reddit, and OpenAI. We're backed by the founders of Clever, ClassDojo, and Coursedog. Our engineering team includes olympiad medalists and hackers from around the world.Our expertise is in building software, so we collaborate closely with schools, educators, and experienced assessors to determine what to build. The schools we work with are collaborators, not customers. We're not another EdTech company trying to profit from the education system. We're here to make a real difference.MissionFix education. We're starting with assessment, one of the most impactful yet burdensome parts of teaching, and expanding from there.VisionA world where students love school and teachers have the time to teach. Where education actually helps the next generation live productive, fulfilling lives. --- Our Research.State of the Art Small-Scale LearningQWK measures agreement between two graders, accounting for the magnitude of disagreements. Being off by one point is much better than being off by three. QWK also considers whether agreement is meaningful or coincidental: if 70% of essays have been graded 4/6, QWK would be near zero for a model predicting 4/6 100% of the time.The chart compares performance on the AES 2.0 Kaggle Competition. The winning solution achieved 0.84 QWK after training on 1,700+ essays. Edexia achieved 0.81 QWK while training on only 20 essays.85× Less Training DataTraditional AI grading systems require thousands of pre-graded essays to train. Edexia's approach achieves comparable accuracy with just 20 examples, making AI grading practical for individual teachers and small schools.1.000.900.800.700.600.840.81Standard ML1,700 essaysEdexia20 essaysOur BackgroundOur team brings research experience from leading institutions in education, machine learning, and assessment science.Harvard UniversityUniversity of CambridgeUniversity of QueenslandUniversity of Technology SydneyInternational Olympiad in InformaticsSmall-Scale LearningConventional AI systems require hundreds to thousands of datapoints to achieve reliability. But that volume of data rarely exists at the class or school level.To work in real educational settings, AI grading systems must adapt from as few as 5 samples and become reliable after 50.November 2, 2025Comparison is KeySaying ‘Essay A is better than Essay B’ is easier than assigning exact grades, for both humans and machines. This unlocks higher reliability when training data is scarce.Read articleFebruary 27, 2026Teaching AI Like We Teach HumansTeachers learn to grade from a handful of exemplars, discussion, and ongoing calibration. LLMs can do the same.Read articleHuman Versus MachineAll machine learning systems require human data as the ‘ground truth’ for training and evaluation. But what happens when that ground truth is flawed?When humans themselves disagree, the definition of ‘accuracy’ in grading becomes far less straightforward.February 27, 2026Taking Each At Their BestUnder ideal conditions, expert human raters can be found to reach 0.95 QWK, yet on other datasets, modern systems are now exceeding the inter-rater reliability of humans.Read articleNovember 11, 2025The Reliability of Human JudgementWhen two trained raters disagree on 35% of essays, which dataset should AI learn from — and how do we know if either rater was consistent?Read article