Skip to content
Video Lab

Lab sessions, on camera

Engineering setups, failure postmortems, and practical AI systems recorded at 4:5 (60 fps). Every episode includes a copy-ready companion prompt or script.

Explore 36 Everyday Hacks & Prompts
Filter Lab:
  • EP 06 60s · 60 FPS (4:5)

    Home NOC Lab with MikroTik: WireGuard & VLAN Bench

    Click to watch session
    2026-11-13MikroTik
    Lab Session · Episode 062026-11-13

    Home NOC Lab with MikroTik: WireGuard & VLAN Bench

    Real routing hardware, sub-millisecond pings, and rock-solid uptime.

    Setting up a residential Network Operations Center (NOC) bench using MikroTik RouterOS 7: isolated guest/IoT VLANs, WireGuard mesh VPN, and automated latency monitoring.

    • MikroTik
    • WireGuard
    • VLAN
    • NOC
    • Networking
    Run Tool ↗
    Watch across:
  • EP 05 60s · 60 FPS (4:5)

    Ubuntu Dev Box from Zero to Repeatable Station

    Click to watch session
    2026-11-13Ubuntu
    Lab Session · Episode 052026-11-13

    Ubuntu Dev Box from Zero to Repeatable Station

    A repeatable Linux environment setup in under 15 minutes.

    A complete walkthrough from fresh Ubuntu install to a hardened backend workstation: user permissions, SSH keys, UFW firewall, Docker daemon, and NVIDIA PRIME GPU offload.

    • Ubuntu
    • Linux
    • SSH
    • Docker
    • NVIDIA
    Run Tool ↗
    Watch across:
  • EP 04 60s · 60 FPS (4:5)

    3-Layer AI Architecture: Cloud vs Local vs Offline

    Click to watch session
    2026-11-11Ollama
    Lab Session · Episode 042026-11-11

    3-Layer AI Architecture: Cloud vs Local vs Offline

    Zero downtime and zero reliance on paid API tokens.

    Inside the Omnidesk AI cascade. How to route tasks intelligently: Cloud APIs (Gemini/OpenAI) for reasoning → Local Ollama (qwen2.5) on GPU → Rule-based offline heuristics when disconnected.

    • Ollama
    • FastAPI
    • Hybrid AI
    • Omnidesk
    Run Tool ↗
    Watch across:
  • EP 03 60s · 60 FPS (4:5)

    R-C-O Prompt Architecture for Production Systems

    Click to watch session
    2026-11-12Prompt Engineering
    Lab Session · Episode 032026-11-12

    R-C-O Prompt Architecture for Production Systems

    The engineering blueprint that makes AI output 100% deterministic code.

    Moving beyond chat prompts. How Senior Engineers use Role, Context, Constraints, and Output (R-C-O) to reliably generate production SQL, Docker, and systemd units.

    • Prompt Engineering
    • DevOps
    • PostgreSQL
    • Systems
    Run Tool ↗
    Watch across:
  • EP 02 54s · 60 FPS (4:5)

    Vision AI Receipt & Business Expense Scanner

    Click to watch session
    2026-11-12Vision AI
    Lab Session · Episode 022026-11-12

    Vision AI Receipt & Business Expense Scanner

    Never lose money on miscalculated paper memos and retail receipts again.

    How to scan 20 crumpled paper receipts into Google Sheets with zero human calculation errors using Multimodal Vision AI and a 3-task error-shielding prompt.

    • Vision AI
    • Google Sheets
    • Retail Tech
    • Aiyaar
    Run Tool ↗
    Watch across:
  • EP 01 60s · 60 FPS (4:5)

    20-Page PDF to Action Plan in 10s

    Click to watch session
    2026-11-11NotebookLM
    Lab Session · Episode 012026-11-11

    20-Page PDF to Action Plan in 10s

    Stop reading 20-page documents linearly — turn AI into a structured workflow engine.

    A 3-step practical AI workflow using Google NotebookLM and Claude Artifacts. Upload raw sources, run 1 precision command, and get a visual map, summary, and action plan 20× faster.

    • NotebookLM
    • Claude Artifacts
    • Local AI
    • Productivity
    Run Tool ↗
    Watch across: