Code.
Diagnose. Repeat.

Software Engineer at Eneba. Building things that matter, one commit at a time.

VIEW MY WORK arrow_forward

My Toolbox

/ 01

Featured Projects

Live projects and works in progress.

biotech
2026

Odis Diagnostic Intelligence

Dermatological diagnostic decision-support system using interpretable deep learning. Classifies skin lesions with case-based reasoning - showing why a diagnosis was made, not just a black-box score.

Next.js TypeScript PyTorch
UI Preview (for now) arrow_forward
timer
v2.0.0

timer

Countdown, stopwatch, and pomodoro in a single 883K static binary with no runtime dependencies.

Rust serde CLI
View on GitHub arrow_forward

/ 02

Tech Stack

The tools and technologies I reach for to ship reliable, performant software.

code

Frontend

TypeScript JavaScript React Tailwind Next.js
dns

Backend

FastAPI Express Symfony PostgreSQL REST APIs GraphQL
neurology

ML & Tools

PyTorch TensorFlow OpenCV scikit-learn

/ 03

Latest Writing

Published research, occasional thoughts, and more writing on the way.

Conference Paper · ICIST 2026

Interpretable Classification of Skin Lesions using ProtoPNet and Expert Feature Integration

Most medical AI tells you what it thinks - not why. Our paper trains a prototype-based classifier on HAM10000 to do both: matching state-of-the-art accuracy on seven lesion categories while exposing its decision logic in terms dermatologists can actually audit. Built on ProtoPNet, anchored to clinical features from the ISIC 2018 dataset. Final version in progress.

check_circle Accepted at ICIST 2026
Coming Soon

More Writing on the Way

I'm working on more articles about software engineering, system design, and the things I'm learning. Stay tuned.

Coming Soon hourglass_empty

/ 04

About Me

Builder, lifelong learner.

I studied Software Systems at Kaunas University of Technology, passionate about building software that makes a difference. My journey started with curiosity about how things work under the hood - and it's led me from backend engineering to full-stack development.

I was a backend engineering intern at Lithuanian Energy Institute (LEI), where I worked with FastAPI to build internal systems. Now I'm a Software Engineer at Eneba, working across the stack with modern tools and technologies.

When I'm not coding, I'm exploring machine learning with TensorFlow and OpenCV, tinkering with side projects, or diving into a good book. Always learning, always building.

Experience

Software Engineer

Eneba · Present

Working across the stack on real-world systems. Building features, fixing bugs, and learning production engineering.

Backend Engineering Intern

Lithuanian Energy Institute (LEI) · 2026

Built internal backend systems with FastAPI. Gained hands-on experience with Python web development and API design.

BSc in Software Systems

Kaunas University of Technology · 2022 - 2026

Thesis: "Explainable Deep Learning System for Skin Lesion Classification" - grade: 10. Studied software systems, built projects, and explored machine learning, web development, and software engineering.

Quick Facts

LocationKaunas, Lithuania
UniversityKaunas University of Technology
FocusFull-Stack & ML
InterestsCode, Books, Tech
Open SourceGitHub

Get In Touch

Always open to new opportunities and interesting conversations.

mail mariusvarna7@gmail.com MariusVarna marius-varna