Selected work / LiftEd

LiftEd

AI for school assessment: checking work, finding gaps, and giving teachers and school leaders a clearer picture of what to do next.

Domain

EdTech

Users

Teachers · Students · Parents · Admins

Core

AI assessment engine

LiftEd

Work checking, gap diagnostics, knowledge map, adaptive practice, school analytics, and role-based product flows.

The work

LiftEd turns assessment into a chain of actions: checking work, diagnosing gaps, mapping progress, and giving the right next step to the teacher, student, parent, or administrator.

What mattered

The hard part was not generating more content. It was making checked work useful: understandable for teachers, fair for students, and concrete enough for school decisions.

What shipped

We built the checking flow, AI-assisted review, adaptive practice, gap diagnostics, role dashboards, and the product rules around attempts, reports, and school-level analytics.

Case study

LiftEd

<a href="https://lifted.kz/" target="_blank" rel="noopener noreferrer">LiftEd</a> helps schools turn checked work into a clear picture of knowledge: what students understand, where the gaps are, and what should happen next. It is not an LMS, not a content library, and not an AI tutor pasted onto the old workflow.

Problem

Marks alone do not explain what to fix.

Teachers spend hours preparing tasks, checking work, and turning marks into something useful. Open answers are the hardest part: they carry more signal than multiple choice, but they are slow to review at classroom scale. For students, feedback often arrives too late. For school leaders, the gradebook shows results, but not the map behind them: which topic fell through, which class needs support, and where the next intervention should start.

Solution

We built the layer between checking and decisions.

The product flow starts with the teacher: create a task, adapt it, launch it in class, and review results. AI helps with the heavy checking work, but the teacher keeps the final judgement and context. After that, LiftEd turns the result into something the school can act on: topic progress, weak spots, adaptive practice, parent visibility, and dashboards for administrators. The core loop is checking → diagnostics → gap map → next action.

Technical work

AI is useful only when the workflow around it is strict.

We treated AI grading as one part of a stricter assessment system, not as a magic box. AI output is connected to rubrics, attempts, teacher review, and visible reasoning, so the result can be inspected and corrected. The surrounding product logic matters just as much: retries, topic gates, adaptive practice, classroom launch flows, parent reports, and analytics that roll up from student to class to school.

Related work

Building AI for a workflow that has to be trusted?

We help turn the AI part into a product people can use every day: clear rules, practical interfaces, review flows, and backend logic that keeps the system honest.

Brief (optional)