All Projects
Mindivo

EduLearn

Adaptive learning platform with personalized content paths.

Client

EduLearn Global

Duration

8 months

Year

2023

Role

Full-stack development, ML pipeline, content system

Overview

An AI-powered education platform that continuously adapts lesson sequencing, difficulty, and format to each learner's performance profile — dramatically improving completion rates and knowledge retention.

The Challenge

A one-size-fits-all curriculum left 60% of learners either under-challenged or overwhelmed. Course completion hovered at 22% and learner satisfaction scores were declining despite new content investments.

Our Solution

We built an adaptive engine using spaced repetition and item response theory to model each learner's mastery level in real time. Content paths are re-sequenced after every session. A React-based authoring tool lets educators create adaptive content without engineering help.

Results

22% → 71%

Course completion rate

+44%

Knowledge retention

50,000+

Active learners

72

Educator NPS

Tech Stack

ReactFrontend
TypeScriptFrontend
PythonML / Backend
scikit-learnML
PostgreSQLDatabase
RedisCaching
S3Storage
AWSInfrastructure

How We Did It

01

Learner Research

Analysed 2 years of engagement data and interviewed 30 learners to understand exactly where and why drop-off occurred.

02

Adaptive Engine Design

Designed the IRT-based mastery model and spaced repetition scheduler before writing a single line of product UI.

03

Content Authoring Tool

Built the educator-facing authoring system in parallel so content could be tagged and structured for the adaptive engine.

04

Beta & Iteration

Ran a closed beta with 2,000 learners for 6 weeks, tuning model weights before the full 50k-learner launch.