Tools, AI Stack and Hardware I Use
The AI models, developer tools and hardware I actively use as a Senior Software Engineer and Web Technologies Consultant in my daily software development work, in enterprise architectures and in the automation behind ismailhakkieren.com.
🤖AI & Agentic Coding (AI Stack)
The models and tools I use every day for architecture design, multi-file refactoring, autonomous agent workflows and RAG pipelines.
Claude Code & Anthropic Claude (Opus 4.6 / Sonnet 4.6)
Main Coding AgentMy primary tool for terminal-based autonomous engineering, deep codebase analysis, multi-step refactoring and writing tests.
Why: Its large context window and the lowest hallucination rate on complex TypeScript/Next.js architectures.
Claude Code Posts (in Turkish)Cursor IDE & Google Antigravity
AI-First IDEMy everyday development environment, with project-level rules (.cursorrules / GEMINI.md), parallel subagents and instant code completion.
Why: It automates repetitive work with zero approvals thanks to the workspace rules and skills I define myself.
.cursorrules GeneratorModel Context Protocol (MCP) Servers
Agent IntegrationThe standard protocol that connects Supabase/PostgreSQL, GitHub, the file system and custom database schemas directly to coding agents.
Why: Querying the live database schema, validating migrations and making safe tool calls without leaving the IDE.
MCP Config BuilderOpenAI GPT-5.4 / DeepSeek R1 & V3 / Gemini
Multi-Model & RAGEmbedding generation (text-embedding-3-small), structured JSON output (Zod), fast classification and cost/performance comparisons.
Why: Instead of running one giant model for every task, routing between hybrid models based on reasoning, speed and cost.
AI Model Comparison🏛️Production Web & Database Architecture
The technology stack that delivers 100/100 Core Web Vitals and semantic search in high-traffic B2C/B2B applications and on ismailhakkieren.com.
Next.js 16 (App Router) + React 19 + TypeScript 5
Core FrameworkReact Server Components (RSC), streaming SSR, static generation (SSG) and an end-to-end type-safe full-stack web architecture.
Why: Maximum SEO performance and developer speed with zero unnecessary client JavaScript.
Vercel AI BoilerplateNeon Serverless PostgreSQL + pgvector
Database & Vector SearchA hybrid RAG setup that combines relational data with 1536-dimensional vector embeddings in the same serverless Postgres database.
Why: It removes the cost of running a separate vector database and delivers millisecond semantic search with SQL + HNSW indexes.
Postgres & pgvector OptimizerTailwind CSS v4 + Geist Typography
UI & Design SystemDark/light mode built on CSS variables, highly readable typography for technical articles and zero runtime overhead.
Why: Fast component-based prototyping while keeping the design consistent.
AWS S3, CloudFront CDN & Sharp
Media & DeliveryAutomatic AVIF/WebP conversion of images, audio article (TTS MP3) files and global edge caching of backups.
Why: Lowering LCP (Largest Contentful Paint) and keeping media traffic away from the main application server.
⚙️Enterprise Software & Consulting Stack
The technologies I use in the e-commerce, payment systems and quality management (QMS) projects I consult on for companies.
C# & ASP.NET Core 9.0 (Multi-Layered Architecture)
Enterprise BackendB2B payment gateways, ERP integrations, JWT Bearer authentication and high-security financial APIs.
Why: A strong type system at enterprise scale, high concurrency performance and easy long-term maintenance.
Angular (Multi-App Workspace) & RxJS
Enterprise FrontendMulti-module enterprise Quality Management Systems (QMS), complex reactive forms and role-based (RBAC) admin panels.
Why: A standardized architecture and a modular workspace structure for large engineering teams.
GEO (Generative Engine Optimization) & Technical SEO Automation
Search EngineeringAutomatic JSON-LD schema generation, llms.txt & llms-full.txt builders, instant IndexNow notifications and CI/CD SEO quality gates.
Why: Getting content quoted as a primary source both in Google’s SERP and in ChatGPT, Perplexity and Claude searches.
AEO & LLM SEO Auditor💻Hardware, Terminal & Productivity
The physical equipment and shell tools I use while writing code and designing system architecture for hours every day.
Apple MacBook Pro (Apple Silicon M-Series)
Main WorkstationQuiet, uninterrupted performance for Docker containers, local LLM tests and multiple Next.js/Angular builds, thanks to the unified memory architecture.
Why: Battery life, the stability of UNIX-based macOS and build speed.
Dual 4K Monitors + Mechanical Keyboard Setup
Desk SetupThe IDE and terminal agents on the main screen; live preview, architecture diagrams and database/log monitoring panels on the side screen.
Why: Seeing code, documentation and telemetry at the same time without context switching.
zsh + pnpm + tsx + Git Hooks (.githooks)
CLI & AutomationCustom TypeScript scripts that run image processing, narration, vector indexing, SEO validation and deployment with a single command.
Why: A deterministic publishing workflow (Zero-Submit Publish) that brings human error down to zero.
Try the AI & Architecture Tools I Use in Your Browser
I turned the MCP configurations, RAG chunking calculations, model benchmark comparisons and AEO/SEO audits from my daily workflow into interactive tools that anyone can use.