Education Technology
·Singapore·10 monthsAI-Powered Talent Management Platform for an EdTech Company
Built an AI-driven talent management platform that maps employee competencies against job requirements, generates personalised training roadmaps, and automates assessment through AI-generated questions and essay marking. The platform enables organisations to identify skill gaps, track development progress, and make data-driven decisions about talent deployment.
Hours
Role profiling time (was weeks)
Personalised
AI-generated learning roadmaps
Automated
Essay marking with human parity
Challenge
This EdTech company identified a gap in the enterprise learning market: organisations invested heavily in training content but had no systematic way to connect training to actual competency development, job requirements, or career progression. The typical enterprise learning experience was a catalogue of courses with completion tracking — an employee could complete dozens of courses and still lack the specific competencies their role required, because there was no intelligent mapping between learning activities and competency outcomes. The underlying data problem was substantial. Job descriptions were written in inconsistent, narrative formats that didn't map to measurable competencies. Training catalogues were organised by topic rather than by competency outcome. Employee skill profiles were either self-reported (unreliable) or based on manager assessments (subjective and infrequent). Without structured, reliable data connecting roles to competencies to learning to assessment, any attempt at intelligent talent management was built on sand. The assessment problem was equally challenging. Traditional multiple-choice assessments were easy to automate but poor at measuring real competency. Essay-based and practical assessments were better measures but required human evaluation, which was expensive and slow. The company needed AI that could generate contextually appropriate assessment questions for any competency and evaluate free-text responses with accuracy comparable to human assessors.
Approach
Reluvate built the platform in three layers: competency mapping, learning pathway generation, and AI-powered assessment. The competency mapping layer uses natural language processing to parse job descriptions and extract structured competency requirements. It maps these against a standardised competency taxonomy, creating a machine-readable competency profile for every role in the organisation. Employee competency profiles are constructed from a combination of assessment results, learning activity completion, manager evaluations, and work output analysis. The learning pathway engine compares an employee's current competency profile against their role's requirements (or a target role's requirements for career development planning) and generates a prioritised training roadmap. The roadmap sequences learning activities based on prerequisite relationships, urgency (how critical the gap is for current role performance), and learning efficiency (which activities close the most gaps simultaneously). As employees complete activities and assessments, their competency profiles update and the roadmap adjusts dynamically. The AI assessment engine generates questions tailored to specific competencies and proficiency levels. For knowledge-based competencies, it generates multiple-choice and short-answer questions. For analytical and synthesis competencies, it generates essay prompts and scenario-based questions. The essay marking system evaluates responses against rubric criteria using NLP, providing scores and detailed feedback that explains where the response met or fell short of expectations. The marking system was calibrated against human assessors to ensure consistency.
Design Notes
The competency taxonomy was the foundational design decision. Rather than creating a proprietary taxonomy from scratch, Reluvate built on established frameworks (SFIA for technology competencies, broader industry frameworks for other domains) and created an extensible structure that organisations could customise. This approach meant that the system was immediately useful with its default taxonomy while allowing deep customisation for organisations with specific competency models. Change management for talent management platforms is particularly sensitive because employees perceive competency assessment as a potential threat — the system might reveal gaps that could affect their standing. Reluvate designed the employee-facing experience to emphasise growth rather than evaluation: the platform shows employees their development opportunities and learning paths, not their deficiency scores. Manager views provide the gap analysis, but the framing is always about investment in development rather than performance judgment. Exception handling in AI-powered assessment must account for the inherent subjectivity of competency evaluation. The system assigns confidence scores to every assessment evaluation. When confidence is below a threshold — typically for complex essay responses where the AI's evaluation diverges from its training distribution — the assessment is routed to a human evaluator with the AI's preliminary analysis as a starting point. This human-in-the-loop approach ensures that edge cases are handled fairly while the AI handles the high-volume, straightforward assessments autonomously.
Result
The platform enables organisations to create structured competency maps for their entire workforce, identify critical skill gaps at individual and organisational levels, and generate personalised development plans that connect directly to measurable competency outcomes. AI-generated assessments have been validated against human assessors and shown to produce consistent, reliable competency measurements. The EdTech company has deployed the platform across several enterprise clients, with competency mapping reducing the time to create structured role profiles from weeks of consulting engagement to hours of automated analysis.
Related Case Studies
Education
Award-Winning AI-Powered CAD Assessment for National Education
Developed an AI-powered automated assessment system for Computer-Aided Design (CAD) student submissions, replacing hours of manual instructor evaluation with instant, educational feedback. The system won the Ministry of Education (MOE) Innergy Award (Gold) — national recognition for innovation in education — and has been adopted internationally.
Retail & Conservation
Computer Vision Retail Analytics and Wildlife Monitoring
Developed computer vision systems for two distinct domains: retail analytics (footfall tracking, customer heatmaps, and employee activity monitoring across store networks) and animal welfare monitoring (behavioural analysis, temperature tracking, and health anomaly detection). The retail system outperformed a major Chinese AI competitor's solution in accuracy benchmarks.
Financial Technology & Media
AI Content Generation for a Major FinTech Industry Conference
Developed an AI content generation system that transformed roundtable discussion transcripts from a major FinTech industry conference into structured, publication-ready reports. The system processed hours of expert roundtable recordings, extracted key themes and insights, and generated polished reports that were published as official conference outputs.