Devgna Vyas — Systems-First Engineer

About

Building systems that think clearly.

I am a BCA student at CHARUSAT University with a systems-oriented approach to software engineering. My core strengths lie in backend development, networking fundamentals, Linux-based environments, and understanding how data and control flow through real systems.

I use AI tools and agents as productivity multipliers — to accelerate design, exploration, and iteration — while retaining full ownership of architecture, logic, and decision-making.

Cybersecurity is a serious area of interest for me, particularly where it intersects with backend systems, networking, automation, and infrastructure security. I actively practice ethical hacking, vulnerability research, and defensive thinking to build more robust systems.

Education BCA, Computer Science
CHARUSAT · 2023 – 2026

Location Nadiad, Gujarat, India

Focus Backend · Networking · Linux
Cybersecurity · AI Engineering

Certifications Crash Course on Python (Google)
Oracle Java Foundations
Introduction to Java
Tools of the Trade: Linux & SQL (Google)

Selected Work

Projects built with depth.

Featured Project

Same.Energy Android Client

An elegant, high-performance Flutter Android client for same.energy — the AI-powered visual search engine. Built with Glassmorphism design and Riverpod state management.

Challenge

Build a native-feeling Android experience for an AI-powered visual search engine that matches the premium quality of the web platform while handling complex image search workflows, offline support, and secure authentication.

What Was Built

A full-featured Flutter application with visual, text, and hybrid image search. Features include masonry grid layout, curated content feeds, collections & bookmarks, full-screen image viewer with download/share, deep linking, and adaptive theming — all wrapped in a stunning Glassmorphism UI.

Architecture

Clean Architecture with feature-first modules. Riverpod for predictable reactive state management, Dio for robust networking with interceptors and retry logic, GoRouter for declarative type-safe routing, encrypted token management, and clickstream telemetry.

Flutter Dart Riverpod GoRouter Dio Glassmorphism Clean Architecture

View on GitHub ↗

Research Project

HERMES

A hybrid AI/ML network security system designed and deployed on ARM clusters for resource-constrained edge environments.

Challenge

Existing intrusion detection systems rely on GPU-backed deep learning or signature-only approaches — both impractical for edge environments with strict power, memory, and compute budgets.

What Was Built

A two-stage hybrid detection pipeline combining signature filtering with a quantized 3-layer deep neural network. The system runs on ARM cluster nodes, enabling distributed monitoring and cooperative intelligence sharing across the edge.

94.7% Novel attack detection

4.2W Power per node

15 kpps Throughput

89.3% Adversarial robustness

Python Deep Learning ARM Edge Computing Network Security IDS

Research

Published work.

HERMES: Design and Deployment of a Hybrid AI/ML Network Security System on ARM Clusters for Edge Environments

Arpankumar G. Raval & Devgna Vyas

7th International Conference on Soft Computing and Its Engineering Applications (icSoftComp 2025)

📅 December 9–11, 2025 📚 Published by Springer Nature 🏛️ CHARUSAT University

A hybrid intrusion detection system combining signature-based filtering with a quantized deep neural network, deployed on ARM-based edge clusters. Achieves 94.7% detection of novel attacks at just 4.2W per node with 89.3% adversarial robustness — 67% less energy than comparable x86 setups.

Capabilities

What I work with.

Languages

Java
Python
Dart
SQL
JavaScript
HTML / CSS
Bash

Backend & Systems

REST API Design
Object-Oriented Design
Clean Architecture
Database Design
Flask

Networking & Security

TCP/IP & Protocols
Ethical Hacking
Vulnerability Research
Intrusion Detection
Network Analysis
Defensive Architecture

Linux & DevOps

Linux Administration
Shell Scripting
Docker
Git & GitHub
CI/CD Workflows

AI & Applied ML

Deep Learning
Model Quantization
Edge AI Inference
AI-Assisted Dev

Mobile & Frontend

Flutter
Riverpod
Material Design
Responsive Web
UI/UX Sensibilities

Philosophy

"Being weak is nothing to be ashamed of… Staying weak is!"

— Fuegoleon Vermillion

I believe in thinking from first principles — understanding data flow, control paths, and failure modes before writing a single line of code. I value performance, clarity, and maintainability over novelty. Given sufficient time and access, I focus on understanding the problem deeply and then building reliable, maintainable solutions.

Photography

Through my lens.

🐾 Animals

Creatures I adore.

✏️ Sketches

Things I draw.

Writing

Words I keep.

Contact

Let's connect.

I'm always open to discussing new projects, research collaborations, or engineering challenges.

vyasdevgna@gmail.com github.com/vyas-devgna linkedin.com/in/devgna-vyas +91 9510292044