Skip to content
View MatoTeziTanka's full-sized avatar
πŸ’­
#Humility
πŸ’­
#Humility

Block or report MatoTeziTanka

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
MatoTeziTanka/README.md

Hey, I'm Seth πŸ‘‹

πŸŽ–οΈ Veteran | πŸ” Cybersecurity Student | πŸ—οΈ Systems Architect

πŸ“ Rhode Island Β· Evidence Coordinator & Technology Analyst at OSDRI Β· Founder of LightSpeedUp

Python TypeScript Rust C# PostgreSQL Linux Proxmox


I build systems that work when nobody's watching.

I don't try to sound smart. I simplify things until they make sense, then build from there. Plain language isn't a style choice β€” it's how I test whether an idea is real. Serious ideas came from simple conversations.

⚠️ A note on AI: I use AI tools extensively β€” they're part of how I build, research, and ship. But every idea, decision, and direction is mine. AI is the tool. I'm the builder.


πŸ›€οΈ The Journey

IT career:        Data centers, consulting, infrastructure
May 2025:         Hit the wall. Stepped away.
Summer 2025:      Recovery. Got healthy. Mental health reset.
October 2025:     Enrolled in cybersecurity. Made the pivot into AI.
4 months in:      Self-hosted governance-native AI stack in production.
Now:              Still building. Helping veterans. Filing patents.

πŸ’‘ Recovery taught me that systems need guardrails β€” including AI systems.


πŸ›οΈ The Stack

These aren't separate projects. They're organs of the same system.

⚑ OLYMPUS β€” Governance-native AI infrastructure. Self-hosted on a Dell R740 (472 GB RAM, 51 TB storage). 26 containers. Fail-closed by design β€” every decision logged, every mutation gated, every service accountable. The constraints are the architecture.

🧠 STYX β€” State-first context engine (patent pending). Compresses meaning instead of storing raw conversation. Benchmarked against RAG and GraphRAG with published methodology β€” results speak for themselves. The algorithm is protected; the evidence is public.

πŸŽ–οΈ VALOR β€” AI-assisted veteran claims platform, built at OSDRI. This is where the stack meets real people. Evidence gathering, statement drafting, intake automation, eFolder analysis. Real cases, real outcomes, daily caseload.

🌱 Sprout β€” A systems programming language. Growing runtime, grammar, and toolchain. Where this goes next isn't public yet.

πŸ“ˆ KERDOS β€” Algorithmic trading infrastructure. Dedicated execution plane in C#/.NET 8. Safety-first paper trading with multi-check preflight.

🏟️ Agora β€” Community companion for OpenAI's Parameter Golf. Live leaderboard with code-based technique classification β€” inspects actual submissions, not PR descriptions.


🧭 How I Think

  • πŸ’¬ Plain language over jargon. If I can't explain it simply, I don't understand it yet.
  • πŸ”¬ First principles over convention. If the answer is "that's how everyone does it," I ask why.
  • πŸ—οΈ Governance is the foundation. Not a layer added later. The system should explain itself at any point.
  • βœ… Verify before claiming. Every assertion should carry evidence.

🎲 Random Facts

  • πŸŽ–οΈ U.S. Military Veteran
  • πŸ“š Cybersecurity student
  • 🧠 Self-taught systems architect
  • β˜• Powered by coffee and spite
  • πŸ”οΈ Building for Titan, not Mars

πŸ”— Links

LinkedIn


"Systems that can't justify themselves shouldn't be trusted to act."

Pinned Loading

  1. styx-results styx-results Public

    STYX: State-first context engine. 3.7x more efficient than RAG on real-world data. Patent pending.