Applying AI in Education: Modernising an LMS for Smarter Online Assessment

A training organisation operating across several markets relied on two disconnected systems for learning management and assessments. Tech JDI integrated both platforms and introduced AI to speed up grading, identify learning gaps, and recommend personalised learning paths, while keeping all certification decisions in the hands of educators.

Client Background

Our client provides competency-based upskilling programmes and certification courses for individuals and corporate clients. Their offerings range from employee onboarding to compliance training, including phishing awareness, serving approximately 25,000 active learners across four countries.

As the organisation grew, they moved from paper-based exams to an online assessment platform that operated separately from their learning management system (LMS). For our client, delivering timely, fair, and well-documented assessment results is essential to maintaining their competitive edge.

Challenges

Moving exams online solved the logistical challenges, but not the workload. As enrolments grew, the gaps between the two systems became the organisation’s biggest bottleneck.

In other words, they wanted to modernise their existing infrastructure by integrating the two systems, reducing manual work, and giving every learner a clear next step after each assessment.

Our Solution: Enhancing the E-Learning Management System with AI-Powered Online Assessment

After reviewing several potential partners, our client chose Tech JDI to modernise their existing platforms without rebuilding them from scratch.

The goal was to reduce repetitive work for educators and give learners faster, more useful feedback, while keeping certification decisions in human hands. AI generates preliminary scores and recommendations, but educators review and approve all results that determine whether a learner passes.

To achieve these goals, we delivered five capabilities aligned with the client’s three priorities. In Phase 1, Capabilities 1–4 and the web dashboard were piloted with one certification programme. After the client approved the pilot, they were rolled out across all programmes and countries.

In Phase 2, we introduced personalised learning recommendations and mobile apps, once enough competency-tagged assessment results had been collected to support relevant recommendations.

1. Centralised Assessment Management

The enhanced platform brings the entire assessment lifecycle into a single, structured system. Administrators can now design, schedule, and monitor assessments from one place, simplifying coordination and reducing operational complexity as the learner base grows.

2. Flexible Question and Scoring Configuration

To support different learning formats and certification standards, the system allows administrators to configure various question types, including multiple-choice, multiple-select, numeric-response, and written-response questions. Scoring rules can also be customised to align with specific course objectives and assessment requirements.

3. AI-Assisted Grading and Faster Results

We introduced an AI-assisted grading system that streamlines assessment evaluation, reduces manual workload, and promotes greater scoring consistency. 

Objective questions are scored instantly using predefined rules, while written responses are evaluated against course-specific rubrics to generate provisional scores, criterion-based justifications, and contextual learner feedback. 

Built-in validation checks flag potentially problematic responses and inconsistent scores for further review, with additional safeguards applied to results near the pass mark. For certification assessments, educators retain final approval over every result.

Learners receive faster, more actionable feedback, while educators spend less time on repetitive grading tasks and more time supporting learning outcomes. 

4. Clear Performance Insights

Rather than presenting only final scores, the platform combines competency-level assessment data with AI-generated insights to reveal deeper performance patterns and learning trends. Educators can track individual learner progress, compare cohort performance, and identify recurring skill gaps across a course. 

AI-generated summaries also highlight questions that many learners answer incorrectly, helping educators pinpoint potential gaps in learning materials or assessment design and make more informed decisions about where improvements are needed.

5. Personalised Learning Recommendations (Phase 2)

Phase 2 started with a rule-based recommendation system as a practical alternative to a more complex recommendation model. 

Each competency-level performance band, for example, a score below 60% in identifying phishing links, is mapped to specific lessons, practice exercises, or resources from the organisation’s approved learning catalogue.

Educators manage these rules and can customise them for each programme. New learners initially receive recommendations based on predefined programme settings until sufficient assessment data is available. Over time, these rules also help build the dataset needed to evaluate and develop a more advanced recommendation model.

How We Implemented It

The Result

Our client kept the systems their teams and learners were already familiar with, while Tech JDI connected both platforms into one integrated solution.

The result is a faster, more efficient assessment process. Objective assessments now return results as soon as the exam window closes, administrators spend less time coordinating exam cycles, and markers receive AI-suggested scores to support the evaluation of written responses. 

Since Phase 2, learners have also received personalised recommendations on what to learn next after each assessment.

The impact in the first three months after full rollout, compared with the three months before implementation:

AI in education isn't just about automating assessments. It's about reducing administrative work, giving learners faster feedback, and helping educators focus on evaluating progress and making informed certification decisions.

For this project, Tech JDI connected the client's learning management and assessment platforms, then introduced AI to support grading, identify learning gaps, and recommend personalised next steps, while keeping educators in control of all certification decisions.

The same principle applies beyond education: start with the workflow you want to improve, then integrate AI where it can make a meaningful difference without replacing the expertise that matters.

Looking to Make Your Existing Systems Work Smarter with AI?

You may not need to replace your existing software to improve efficiency. Tech JDI helps businesses connect disconnected systems, automate repetitive tasks, and integrate AI into existing workflows, with solutions tailored to their needs.

Frequently Asked Questions