Welcome to my GitHub profile! I'm a software developer with 4+ years of professional experience, focused on building scalable web applications, developer tools, AI-powered systems, and modern user interfaces.

My main background is in front-end engineering, but over time I’ve expanded into AI agents, LLM infrastructure, desktop applications, backend systems, and low-level programming.

• Front-end: JavaScript, TypeScript, React, Lit 3, Web Components, Shadow DOM, HTML, CSS/SCSS, Vite, Rollup, Nunjucks, TanStack Query
• Web Components & Design Systems: Custom Elements, Lit, component architecture, design tokens, CEM (custom-elements.json), reusable component libraries
• Back-end: Node.js, Python, REST APIs, PostgreSQL, Supabase, SQLite, Edge Functions
• Testing & Quality: Vitest, Playwright, Open WC Testing, integration testing, regression testing, ESLint
• Desktop & Cross-platform: Tauri, Electron, WebView-based applications
• AI & LLMs: AI agents, multi-agent systems, RAG, MCP, tool calling, skills, structured tools, agent memory, model routing and AI-assisted development
• Local AI & Inference: Ollama, open-weight LLMs, local inference, multimodal models, SGLang, OpenAI-compatible endpoints
• AI Infrastructure: GPU inference, serverless GPU workloads, Runpod, Docker, model serving and hybrid local/cloud AI architectures
• Systems & Low-level: C, C++, Linux, Arch Linux, GPU experimentation and emulator development
• Developer Tooling: Git, GitHub, npm, Vite, Rollup, esbuild, Babel, VS Code, Docker

A major area of my current work and research is agentic software development.

I experiment with architectures where AI models can use tools, MCP servers, skills, RAG, memory and specialized subagents to solve complex software engineering tasks.

Some areas I actively explore include:

• Multi-agent orchestration and task delegation
• Model routing based on task complexity
• MCP tools and servers
• Agent skills and reusable workflows
• Retrieval-Augmented Generation (RAG)
• Long-term and contextual agent memory
• Tool calling and structured outputs
• Local and cloud LLM inference
• Open-weight models
• AI coding agents
• Hybrid local/cloud AI systems
• GPU-backed model serving

I'm particularly interested in building AI systems where models are not just chat interfaces, but software agents capable of researching, planning, using tools, modifying code, running tests and validating their own work.

My strongest area is building maintainable and scalable front-end architectures.

I work extensively with React, Lit and native Web Components, including component communication, Shadow DOM, design systems, responsive interfaces and reusable UI architectures.

I also work with modern build tooling such as Vite, Rollup and esbuild, with a strong focus on:

• Performance
• Accessibility
• Responsive design
• Component architecture
• Maintainability
• Internationalization
• Developer experience
• Integration and regression testing

I enjoy working beyond traditional web development.

My personal projects involve areas such as:

• Emulator architecture and reverse engineering
• C and C++ development
• GPU programming and experimentation
• Linux and Arch Linux
• Desktop applications with Tauri and Electron
• Local LLM inference
• AI development tools
• Visual development tools
• Canvas-based editors
• Developer tooling and automation

I enjoy understanding how systems work beneath the abstraction layer and applying that knowledge back to web and AI development.

I'm especially interested in projects involving the intersection of:

Web Development + AI Agents + Developer Tools + Systems Programming

My goal is to build software that is not only functional, but also maintainable, performant, extensible and capable of integrating increasingly intelligent workflows.

Feel free to explore my repositories and experiments.