Skip to main content

About Masar AI

Your path, guided by data.

Problem statement

The core problem is that traditional career counseling is often inaccessible, expensive, and non-personalized, leading many individuals to make uninformed decisions that result in career dissatisfaction and skill mismatches. Conventional methods rely on static aptitude tests and limited human resources, which fail to keep pace with a rapidly evolving job market and the emergence of new technical roles. Consequently, there is a critical need for a scalable, data-driven AI system that can analyze vast amounts of real-time industry data and individual user profiles to provide objective, personalized, and actionable career pathways.

Abstract

Many students face difficulty in selecting appropriate career paths due to limited access to personalized and reliable guidance. Traditional career counseling methods are often generalized, inconsistent, or unavailable to all learners, resulting in decisions based on uncertainty rather than informed self-assessment. Consequently, students may choose academic subjects or career directions that do not align with their abilities, interests, or personality traits, leading to long-term academic dissatisfaction and professional challenges.

The proposed AI Career Guidance System addresses this issue by providing a data-driven, accessible, and personalized platform for career planning. The system leverages machine learning techniques to analyze key student data, including academic performance, personal interests, and personality profiles. Based on this analysis, it generates tailored career recommendations that align with each student's unique strengths and preferences. Additionally, the system identifies skill gaps and suggests relevant courses and learning pathways to support students in achieving their career goals.

By offering continuous, individualized support, the AI Career Guidance System enhances decision-making and reduces reliance on guesswork. It ensures that students receive accurate and timely guidance, empowering them to make informed academic and professional choices. Ultimately, this approach contributes to improved educational outcomes and better alignment between students' potential and their future careers.

How the system works

Masar combines 2 models rather than relying on one, because neither answers the whole question on its own.

  1. 1. Sector classifier

    A Random Forest trained on 5,000 generated student profiles predicts which of the 15 sectors fits a student. Logistic Regression and K-Nearest Neighbours were trained as baselines for comparison.

  2. 2. Similarity layer

    Each career has an ideal profile vector. We compare the student's 20-dimensional profile against it with a mean-centred cosine similarity, which ranks individual careers inside and across sectors.

  3. 3. Weighted blend

    The final score is 45% classifier, 45% similarity and 10% labour-market demand. Demand is capped deliberately: a guidance system that chases fashion would recreate the problem it exists to solve.

  4. 4. Explanation

    For each match we surface the 3 profile dimensions that contributed most, and only where the student and the career are both strong — a shared strength, never a shared absence.

A note on the data

No public dataset links UAE school grades, RIASEC interests and Big Five traits to career outcomes. Rather than claim one, Masar ships a deterministic synthetic dataset generator with a documented causal story, and states its limits openly. High accuracy on this data demonstrates that the pipeline works — not that the recommendations are externally validated.

Team

  • Saif Qais AhmedGroup Leader & Frontend
  • Khaled Mohammed AlMemariCo-Leader & Database & DevOps
  • Mansour Buti ALShmasiQA & Documentation
  • Mubarak Awad AlAmroBackend & API
  • Zayed Saif AlBlooshiML Engineer
Supervisor
Hamdy Hersi
Institution
Applied Technology School — Al Ain
Academic year
2026