Third-year software engineering student at NTNU Trondheim. I enjoy building practical and scalable solutions, with a special interest in web development and AI.

My bachelor's thesis is written in collaboration with Piscada AS in Trondheim. The project explores how retrieval-augmented generation (RAG) and related tooling can be used in practice, combining theory with concrete implementation. The work focuses on an AI assistant that explains energy data in natural language and provides targeted energy-saving advice for commercial buildings.

The thesis has directly shaped the AboutMe portfolio below — API design, document pipelines, vector storage, prompt versioning, and evaluation of language models all stem from the research and development carried out during the project.

📺 Video about the project: Watch on YouTube

AboutMe (live portfolio): A document-grounded AI chat (RAG) trained on my background and projects. The UI supports Norwegian and English. Stack: Vue 3, Spring Boot, and PostgreSQL with pgvector (relational data and embeddings in one database). Features include an admin document pipeline, prompt versioning, an experiment/evaluation framework, and optional PostHog analytics.
Live: kevindmazali.me · Repo: github.com/kdm-kev-NTNU/AboutMe

Krisefikser (IDATT2106 team project): Crisis preparedness platform built in a 3-week sprint — maps, shelters, news, households, and resources — with Vue 3.5 + Tailwind 4 on the front end and a Spring Boot 3.4 + JWT + MySQL backend. CI with Vitest/Cypress and JaCoCo coverage ≥80%.
Repo: github.com/idatt2106-v25-02/krisefikser

Languages & Frameworks	Databases & Tools	AI & Learning

• Bachelor's Thesis — Piscada AS (IDATT2901, Spring 2026)

  • Exploring RAG (retrieval-augmented generation) in an industrial context with Piscada AS in Trondheim.
  • Built the AboutMe portfolio as the practical implementation: document pipelines, vector storage (pgvector), prompt versioning, and evaluation.
  • 📺 Video about the project

• Krisefikser (Crisis Platform) (IDATT2106, Spring 2025)

  • Built a crisis preparedness platform with Vue 3.5 + Tailwind 4 frontend and Spring Boot 3.4 + JWT + MySQL backend.
  • Delivered maps, shelters, news, household coordination, and resources; CI with Vitest/Cypress and JaCoCo coverage ≥80%.

• Content Student Assistant @ Engage Resource Hub (NTNU) (06/2025 – 12/2025)

  • Managing and publishing educational content for a cross-program resource platform.
  • Supporting feature development and assisting with technical and design improvements.
  • Collaborating with multiple study programs to make digital resources accessible and user-friendly.

• Student Assistant @ NTNU (Programming, Numerics, and Security — INGT1002) (08/2025 – 12/2025)

  • Supporting students with Python programming, numerical computations, and ICT security fundamentals.
  • Assisting with the design and evaluation of weekly exercises and final assessments.
  • Providing guidance to help students apply programming and numerical methods to real-world engineering problems.

• LEGO Data Analysis Project

  • Regression, clustering, and classification on LEGO datasets using Python.
  • Explored pricing differences between LEGO sets for different demographics.
  • Repo: LEGO Analysis

• Roadmap (Algorithms & Data Structures)

  • Implemented shortest-path solutions using Dijkstra and ALT search algorithms on Nordic map data.
  • Added functionality to locate nearest points of interest (POIs).
  • Repo: Roadmap