I tune Linux systems until they scream — in a good way. And I help creators leverage AI tools without paying subscription fees.

What I Do

eGPU / Oculink Performance Tuning Your external GPU shouldn’t feel external. My open-source tuning engine locks PCIe links, disables mobile throttling assumptions, and stabilizes the connection so your docked GPU runs like it’s soldered to the motherboard.

Requires a Linux distribution with systemd (Arch, CachyOS, Fedora, Ubuntu, Debian, etc.). Windows is not supported.

The script is free. Download it, read it, run it. It works on any SFF mini PC + Oculink eGPU setup. Windows is not supported.

If you want me to run it for you via remote session, walk you through what each line does, or troubleshoot if something goes sideways, that’s when I charge. Starting at $75 for a remote tune session + 30 days of follow-up.

Mini PC & SFF Optimization Small form factor doesn’t mean small performance. Kernel scheduler tuning, thermal management, LUKS storage optimization, and gaming frame-time stability. The knowledge is free, my time isn’t. Starting at $50 for a remote session.

AI Content Creation — Local, Private, No Subscriptions

  • AI Video Generation — Text-to-video, image-to-video with Google Veo, Stable Diffusion animations. I can generate custom video content for your brand, social media, or creative projects.
  • AI Image Generation — Stable Diffusion fine-tuned to your style, character consistency, product visualization, concept art.
  • AI Storybooks & PDFs — Illustrated children’s books, technical guides, visual narratives. Generated locally, fully customizable.

Pricing varies by complexity — email me with your project.

Technical Writing & Blog Content I write about Linux performance, AI tooling, and creative tech. Whether you need a tutorial, a deep-dive benchmark article, or a script for your YouTube channel, I can deliver content backed by real expertise.

Starting at $200 for a 2,000-word technical article with benchmarks and custom tooling.

How It Works

  1. You send me your specs — lspci, uname -a, what you actually do with the machine (or your creative brief)
  2. I build a custom solution — not generic advice, your specific situation
  3. You test it — I include a rollback path, nothing is permanent until you say so
  4. I follow up — 30 days of tweaks if something isn’t quite right

Why Me

  • I run the exact setups I tune — CachyOS, SFF mini PC, Oculink eGPU
  • My tuning engine is open source and battle-tested
  • I generate AI content locally on real hardware, not through paid APIs
  • I’m not a company — you talk to the person who actually wrote the code

Get In Touch

Email: tattvamasi68@gmail.com
GitHub: scottyad
Response time: Usually same day, always within 24 hours

Not sure if your problem is something I can fix? Email me anyway — if I can’t help, I’ll tell you who can.