Computer
MacBook Pro 16-inch, M-series Max, 64 GB unified memory
Starter specification for sustained architecture and local AI workloads.
Workstation
A low-friction environment for architecture, code, and local AI work.
The setup is optimized for deep technical work across large diagrams, multi-repository delivery, local containers, and model-assisted workflows.
Operating posture
macOS-first with Windows and Linux test environments
Core workload
Architecture, full-stack development, containers, and local inference
Configuration principle
Keyboard-first, reproducible, and quiet by default
Starter configuration based on the approved portfolio brief. Exact hardware, extensions, and dotfile links should be confirmed by Panagiotis before this page is treated as a purchasing inventory.
01
Primary workstation
High-memory Apple Silicon keeps large workspaces, containers, and local models responsive without turning the desk into a server room.
Storage
2 TB internal SSD + encrypted external NVMe workspace
Active repositories stay local; archives and large models move to dedicated storage.
Secondary environment
Windows 11 workstation / WSL2
Used for Windows-native packaging, desktop utilities, and enterprise compatibility checks.
02
Displays
One high-density primary canvas for diagrams and code, plus a portrait or secondary panel for reference material.
Primary
27-inch 5K or 32-inch 4K display
Text clarity and usable diagram space matter more than extreme refresh rate.
Layout
Centered primary + portrait reference display
Keeps the active system model centered while documentation and logs remain visible.
Window management
Repeatable arrangements for architecture, code review, writing, and incident modes.
03
Terminal & shell
A fast, legible command surface with portable configuration and conservative defaults.
Terminal
Ghostty or WezTerm
GPU-accelerated rendering, native feel, and configuration that can live beside dotfiles.
Shell
Minimal prompt context with Git state, runtime version, and execution status.
Core CLI
ripgrep, fd, fzf, jq, gh, and mise
Small composable tools keep repository and platform navigation fast.
04
Editor & knowledge system
VS Code for broad ecosystem compatibility; Obsidian for durable architecture thinking and local-first knowledge.
IDE
TypeScript, Python, Rust, YAML, containers, Playwright, and GitHub workflow support.
Core extensions
ESLint, Prettier, Ruff, rust-analyzer, YAML, Playwright, GitLens
Automated feedback stays close to the edit loop without hiding repository-level quality gates.
Notes
Obsidian + Dynamic Wide Content
A narrow reading column that can expand diagrams, tables, and code only when the material needs it.
05
Local AI runner
Cloud models for frontier reasoning; local runners for privacy-sensitive experiments, offline evaluation, and latency checks.
Runtime
Local model management, OpenAI-compatible endpoints, and controlled evaluation environments.
Containers
Docker Desktop
Repeatable vector stores, databases, and supporting services for retrieval and agent prototypes.
Operating rule
No production secrets or customer data in experimental prompts
Local execution supplements governance; it does not replace data-classification controls.