Skip to content
KVANT

Research

Software we build, and research.

Our managing director Ajdin Muratović, a PhD candidate at the University of Zenica, publishes research on web application development, data security and artificial intelligence. The papers were presented at international conferences and published in IEEE Xplore.

published research papers
6
published research papers
papers in IEEE Xplore
5
papers in IEEE Xplore
international conferences
4
international conferences
Ajdin Muratović

Ajdin Muratović

Managing Director, KVANT d.o.o.

  • PhD candidate, University of Zenica · software engineering
  • Master of Information Technology and Software Programming
ResearchGate profile

From research to practice

What it means for our clients.

Secure data without a slow system

We know how to protect personal data with encryption while keeping the application fast.

Faster development, fewer errors

Automation and AI in development shorten build time and leave repetitive work to tools.

AI changes under control

When AI changes code in a live system, we measure whether its behaviour changed before users notice.

Publications

Papers

2026

INFOTEH-JAHORINA 2026 · 25th International Symposium · March 2026

Baseline Capture and Drift Index: A Framework for Validating AI-Assisted Changes in Existing Web Systems

Ajdin Muratović et al.

AI tools increasingly change code in systems that are already in production, and every such change risks silently altering how the system behaves. The paper proposes a framework that captures the system’s reference behaviour and measures deviations with a layered Drift Index.

Result

The method reliably separates harmless content changes from serious API contract violations, with far fewer false alerts than simple response comparison.

Paper on IEEE Xplore

DOI: 10.1109/INFOTEH68759.2026.11477738

INFOTEH-JAHORINA 2026 · 25th International Symposium · March 2026

Artificial Intelligence as a New Teaching Tool: Challenges and Dangers

Ajdin Muratović, co-author

The paper examines the benefits and challenges of bringing artificial intelligence into education, and how ready teachers and students are to use it well.

Result

Successful use requires a multidisciplinary approach and a balance between technology and teaching goals.

Paper on IEEE Xplore

DOI: 10.1109/INFOTEH68759.2026.11477666

2025

Spoznaja · Journal of the Faculty of Philosophy, University of Zenica · December 2025

Gamifikacija u Moodle LMS-u uz AI-podržano ocjenjivanje na modelu učenja konvolucijskih neuronskih mreža

Ajdin Muratović et al.

A gamified online course model in Moodle LMS with five progression levels and an AI add-on that grades essays automatically and gives learners instant feedback.

Result

Combining gamification with AI-supported grading increases learner motivation and makes complex concepts easier to understand.

Spoznaja journal

INFOTEH-JAHORINA 2025 · 24th International Symposium · March 2025

Development of a Methodology for Automated Backend Code Optimization Triggered from a Frontend Web Application Using Generative Artificial Intelligence Models

Ajdin Muratović et al.

A methodology in which backend code optimisation is triggered from within the web application and carried out by generative AI models while the system keeps running. It combines Vue.js, Laravel, a PHP parser and the OpenAI API.

Result

Across 20 tested functions: a 70% optimisation success rate, 62.6% shorter execution time and 18.18% fewer lines of code, with unchanged behaviour.

Paper on IEEE Xplore

DOI: 10.1109/INFOTEH64129.2025.10959222

2024

TELFOR 2024 · 32nd Telecommunications Forum, Belgrade · November 2024

A Method for Accelerating Encrypted Data Search in Laravel Applications Using a Coded Search Parameter

Ajdin Muratović et al.

Personal data should be stored encrypted, but then searching is slow, because every record has to be decrypted to check for a match. The paper introduces a Coded Search Parameter (CSP) that narrows down candidates before decryption.

Result

Searching encrypted data is on average 73.58% faster, without weakening data protection.

Paper on IEEE Xplore

DOI: 10.1109/TELFOR63250.2024.10819104

TELFOR 2024 · 32nd Telecommunications Forum, Belgrade · November 2024

Innovative Approach to Accelerating Laravel Backend Generation from Vue.js Components Using Axios, Artificial Intelligence and Artisan Commands

Ajdin Muratović et al.

After building the user interface, developers usually create the matching backend by hand. The paper presents a system that automatically generates Laravel models, migrations and controllers from a Vue.js component, using Axios, Artisan commands and AI.

Result

Initial development time of Laravel applications was reduced by 33.96%, with fewer human errors.

Paper on IEEE Xplore

DOI: 10.1109/TELFOR63250.2024.10819138

Have a project in mind?

Tell us what is slowing you down. The first conversation is free, takes 30 minutes and commits you to nothing.

Book a free consultationFree website proposal
Phone
+387 61 813 300
Email
[email protected]
Office hours
Mon – Fri, 08:30 – 17:00 CET