Architecture that knows no limits.

Senior Symfony development. I build robust backend systems, blazing-fast PWA applications and integrate advanced AI models for maximum automation of your business. 13+ years of experience.

Technology Arsenal

// 01

Symfony Core

8+ years, 30+ enterprise projects. Specialisation in robust backend architectures, high performance and security.

Symfony 8 PHP 8.5 DDD PostgreSQL
// 02

AI Integration

LLM integrations (GPT-4, Claude), RAG systems for corporate data and autonomous agents. AI as a tool for efficiency, not a buzzword.

OpenAI API RAG Python Vector DB
// 03

PWA & State Machines

Blazing-fast frontends with instant response. Offline-first approach and absolute control over application state.

Stimulus Turbo Service Workers XState
// 04

Infrastructure

Modern deployment and scaling. Containerization, CI/CD pipelines and real-time system monitoring.

Docker CI/CD Linux Admin Nginx

Engineering Pillars

Solutions designed for scalability, performance and long-term maintainability.

// Architecture

Robust Backend Systems

Focus on performance, multithreading on Linux/Windows, secure APIs and complex relational structures. Symfony as the core of your business.

// Interactive

Interactive Applications (PWA)

Tools that work like native apps. Emphasis on UX, State Machine architecture and instant response on every device.

// Intelligence

AI & Automation Scripts

Replacing manual work with intelligent algorithms. From data processing to complex real-time decision trees.

13+ Years of engineering
30+ Enterprise Symfony projects
100% Code Quality & Coverage
0 Compromises ("Everything is possible")

6 words

A social platform where a question receives an AI image cover and the community replies in exactly 6 words. Key innovation: Prompt-on-Prompt technique, where AI first translates an abstract query into a precise visual descriptor before it enters the diffusion model. Fully asynchronous pipeline via Symfony Messenger, credits, automatic categorization, typo-tolerant search, and a custom auto-moderation system.

Prompt-on-Prompt Pipeline Two-phase AI transformation: GPT-4 first translates an abstract question into a precise visual descriptor, DALL-E 3 generates a cover at 1792×1024.
Fully Asynchronous AI Pipeline 7 specialized Symfony Messenger handlers — images, answers, categories, nicknames, emails. UI never waits for AI.
  • OpenAI GPT-4o + DALL-E 3
  • Asynchronous Symfony Messenger + Scheduler
  • Prompt-on-Prompt (two-phase image generation)
  • Loupe — typo-tolerant full-text search (PHP-native)
  • Weighted Report Score with dynamic reporterTrust
  • Custom Imagick composite engine (brick grid + sigmoid blend)
  • 203 questions · 1,274 answers · 39 users (Aug 2026)
AiImageGenerateHandler.php
#[AsMessageHandler]
readonly class AiImageGenerateHandler
{
    public function __invoke(AIImageGenerate $message): void
    {
        $question = $this->questionRepo->find(Uuid::fromString($message->questionId));

        // Phase 1: GPT converts abstract question → visual descriptor
        $prompt = $this->promptGenerator->generateImagePrompt(
            $question->getQuestion()
        );

        // Phase 2: DALL-E 3 generates the actual image at 1792×1024
        foreach ($this->generator->generateImage($prompt, 1792, 1024, 1) as $resource) {
            $this->imageService->createImage($resource, $question);
            $this->em->flush();
            return;
        }
    }
}
// Other projects
// Methodology

Architectural Process

01

System Analysis

Deconstruction of the problem into elementary parts. Identification of bottlenecks and definition of uncompromising goals.

Status: Analyzing dependencies...
02

Proof of Concept

Rapid validation of critical algorithms and integration points. Elimination of technological risks at an early stage.

Logic: Validating O(n) complexity...
03

Architecture Blueprint

Design of a robust system with regard to scalability, security and future AI automation.

Stack: Symfony Core + Vector DB
04

Core Engineering

Writing highly efficient code, implementing business logic and ensuring 100% test coverage.

Git: feat(core): implemented async drivers
05

AI & Fine-Tuning

Integration of LLM models and prompt tuning for specific company data. Increasing process efficiency.

AI: RAG engine optimization active
06

Enterprise Ops

Production deployment, real-time monitoring and continuous performance optimisation under load.

Uptime: 99.99% | Latency: 45ms
// References

What clients say

"S pánom Jozefom spolupracujem niekoľko rokov. Či to už bol projekt webovej alebo mobilnej aplikácie, konzultácie, následné úpravy a navrhované zmeny z mojej strany, vždy som sa stretol s profesionalitou, ústretosťou a kreativitou. Pán Jozef je veľký profesionál. Ďakujem za výborne odvedenú robotu."

LJ
PhDr. Ľubomír Jaššo Director, YouCity

"Služby sme využili viackrát, sme veľmi spokojní. Vždy rýchle, ústretové, efektívne a spoľahlivé riešenie našich požiadaviek."

RB
Ing. Radovan Baláži RB IT Service & Consulting GmbH

"Toller Service, Schnelle Abwicklung, Große Professionalität, Angemessenheit. Jozef hat für uns viele spezielle Anforderungen im Bereich der Datenerfassung und -verarbeitung gelöst. Wir sind sehr zufrieden."

BL
Bergbahnen Lungau GmbH & Co KG Austria

"Poskytované služby sú na veľmi vysokej úrovni. Hľadanie riešenia každej požiadavky v zmysle hesla „Všetko sa dá". Určite využijeme služby aj v budúcnosti."

PD
Peter Dolinajec ODIS obchodná, stavebná a konzultačná s.r.o.

"Veľmi dobrá komunikácia, rýchla odozva na riešenie problémov, zodpovedný prístup a výborne poradenstvo."

BR
Buček Róbert vasholubnik.sk
contact.sh

Have a specification? Or just a complex problem?

We are ready for any engineering challenge. We will design a solution that works.