Full Stack Developer & AI Engineer building with LLMs — RAG pipelines, fine-tuning, multi-agent orchestration & AI-driven development. Laravel Ecosystem Expert | AI-Powered DevOps | Cybersecurity.
// System Status
class Engineer {
public $name = "Harun Geçit";
public $role = "AI Engineer";
public $status = "Available";
public function getRoles() {
return [
"Full Stack Developer",
"AI Engineer",
"LLM / RAG Developer",
"DevOps & Security",
];
}
}Try some commands: help, ai, about, skills, experience, contact
I'm a Full Stack Developer & AI Engineer who designs scalable systems and builds with Large Language Models every day. With 15+ years in software and 2+ years of hands-on AI engineering, I also work as a DevOps Engineer, System Administrator, Cybersecurity Specialist, and Software Advisor.
On the AI side, I build production RAG pipelines on PostgreSQL + pgvector, design fine-tuning and PageIndex-based training workflows, and orchestrate multiple LLMs together — managing token budgets, system prompts, custom skills and web-search agents. My daily toolkit is LLM-driven: Claude Code, Codex, Gemini CLI, OpenCode, Commander.ai, Cursor, TRAE and VS Code AI agents.
I'm deeply passionate about the Laravel Ecosystem — from Livewire and Inertia.js to Filament, Forge/Vapor. Backend expertise spans PHP, JavaScript, Go, SQL, and Python. I run AI-driven development, testing, staging and production workflows, with LLM-based code review and AI security scanning wired into my CI/CD pipelines on Docker, Kubernetes, Nginx and AWS/GCP.
Years Coding
Years AI Engineering
AI Tools in Daily Use
Companies
Projects+
2+ years building production AI — RAG, fine-tuning, multi-agent orchestration & an AI-driven software lifecycle. I build with LLMs, every day.
LLM pair-programming & agents writing, refactoring and documenting code
AI-generated test suites, coverage analysis & automated bug triage
LLM-based code review, AI security scanning & release-note generation
AI-assisted CI/CD, anomaly detection & intelligent incident response
I develop primarily through LLM-based CLIs and coding agents — driving features end-to-end from the terminal, from scaffolding to refactor to docs.
Heavy daily use of agentic IDEs and in-editor assistants — running autonomous agents for multi-file edits, reviews and repo-wide tasks.
Production Retrieval-Augmented Generation on PostgreSQL + pgvector — embeddings, chunking, hybrid & semantic search, reranking and PageIndex document indexing.
Orchestrating multiple LLMs on a single task, designing system prompts, building custom skills, and accounting for tokens & cost across providers.
Building web-search skills and autonomous research agents that expand an LLM's source count, fetch live context and ground answers in citations.
Designing fine-tuning and PageIndex-based training pipelines — dataset curation, evaluation harnesses and shaping model behavior to a domain.
Using LLM agents for automated code review, code security analysis, vulnerability detection and secure-by-default remediation across the codebase.
Defining LLM boundaries & determinism, adding guardrails, and optimizing performance, resource consumption and stability of AI systems in production.
LLM-based automation across dev, test, staging and production — generating pipelines, reviewing diffs and triaging failures automatically.
Feel free to reach out for collaborations or just a friendly chat!