How E-Porikkha fits your school.
Three types of schools. Three real challenges Bangladeshi teachers face every exam season. Here's how E-Porikkha helps each one.
From three weekend marathons to three review hours — for hundreds of students.
A Bangla department running term exams for hundreds of students across multiple sections. Before E-Porikkha, four teachers spent three full weekends manually grading literature responses. The same workload now takes three afternoons of review.
Bangla literature questions don't have a single right answer. Students paraphrase, draw on personal interpretation, and use mixed Bangla-English when explaining concepts. Standard OCR couldn't read handwriting; keyword grading couldn't grasp meaning.
Onboarding took under 30 minutes. Answer keys were uploaded as PDFs. The semantic grading model handled paraphrased answers correctly — and surfaced borderline cases for teacher review.
Results went out significantly earlier than the previous semester. Teachers reported reclaiming entire weekends. Inter-marker consistency improved — eliminating most parent disputes about partial credit.
“I used to spend my entire weekend grading. Now I spend a couple of hours on Friday reviewing the AI's suggestions. The accuracy on Bengali answers surprised me — it actually understands what the students meant, not just what they wrote.”
From single-subject pilot to whole-school adoption in one term.
A school started with History. Within six weeks, four more departments asked to join — Bangla, Geography, Civics, and Religious Studies. By term-end, teachers across five subjects were grading on E-Porikkha.
The school principal wanted a system that could span subjects without requiring custom configuration per teacher. Different departments had different grading styles, rubrics, and confidence requirements.
Per-question rubric configuration meant every teacher could keep their own style. Confidence threshold sliders let conservative graders review more, while comfortable graders auto-approved more. One platform, fitted to each teacher's preference.
What started as a History-only pilot became a whole-school standard within a single term. The principal cited faster results, fewer disputes, and dramatically improved teacher satisfaction as the deciding factors.
“Our History department grades hundreds of papers per exam cycle. It used to take weeks. With E-Porikkha it took days — and the consistency between markers improved dramatically. No more disputes about partial credit.”
Trust-first rollout: confidence scoring won over a sceptical staff room.
The college's leadership wanted AI grading — but senior teachers were sceptical. The pilot focused on transparency: every AI mark came with a confidence score, source-text highlight, and full audit trail. Trust was earned, not assumed.
Senior teachers feared losing oversight of grading decisions. The school's reputation depends on grade defensibility — every mark must be justifiable to parents, students, and oversight bodies.
Every AI-assigned mark shows the matched section of the model answer, the confidence score, and the reasoning. Teachers can override any decision with one click — and every change is logged with timestamp and rationale.
Within four weeks, sceptical senior teachers became the strongest internal advocates. Every grade question was answered with documented evidence — giving teachers the tools to handle parent queries with confidence.
“As a school principal, what matters to me is trust. E-Porikkha shows detailed confidence scores for every mark — so when parents ask questions, teachers have clear evidence. That transparency is invaluable.”
Ready to see this in your classroom?
Early access is open for Bangladeshi institutions. Start with free trial credits — no card needed.