Skip to content
View ozlar34's full-sized avatar
🎯
Focusing
🎯
Focusing
0 followers · 10 following

Achievements

Achievement: Pair Extraordinaire

Achievements

Achievement: Pair Extraordinaire

Block or report ozlar34

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
ozlar34/README.md

Oz - Project Manager (AI Operations)

I build the AI systems a team actually works with: automating the busywork, keeping people in the loop where it counts.

I spent 8 years running community operations at scale: 250K+ player community, 131K peak concurrent, 15 broadcast languages at Com2uS. Now I build the AI workflows, evals, and enablement systems that let small teams operate like bigger ones, and prove they actually get used.

Selected work

  • job-match-radar: self-hosted n8n + Supabase pipeline that scrapes listings, scores them against a personal rubric, and emails a weekly digest, with an eval + guardrail harness regression-testing the LLM scorer. Production automation, not a demo.
  • llm-judge-evals: a dependency-free harness for keeping an LLM-as-judge honest: gate tests on documented mis-scores, a human-in-the-loop guardrail, and version-keyed drift detection.
  • claude-code-skills: Claude Code skills I built for my own stack: capture, handoff, and knowledge-ops utilities across Obsidian, Notion, and NotebookLM.
  • second-brain-system: a self-documenting knowledge + automation system: Obsidian as the canonical store, Claude Code + local LLMs as the activation layer, drift tripwires as guardrails.

Stack

n8n Supabase Claude Code Ollama Codex Antigravity

Contact

Email LinkedIn

Last updated: June 2026

Pinned Loading

  1. job-match-radar job-match-radar Public

    Self-hosted n8n + Supabase pipeline that scrapes LinkedIn and a watchlist of company ATS endpoints, scores listings against a personal rubric, emails a weekly digest, and regression-tests the LLM s…

    Python 1

  2. llm-judge-evals llm-judge-evals Public

    A dependency-free harness for keeping an LLM-as-judge honest: gate tests on documented mis-scores, a human-in-the-loop guardrail, and version-keyed drift detection.

    Python

  3. job-application-engine job-application-engine Public

    Agent-driven system for crafting defensible, tailored CVs & cover letters — Claude Code skills with traceable fact pools, JD-strand mapping, and hard gates. Synthetic worked example.

    HTML

  4. claude-code-skills claude-code-skills Public

    Claude Code skills I built for my own stack — Obsidian, Notion, capture/handoff, and session workflow utilities.

    Python 2

  5. second-brain-system second-brain-system Public

    A designed, self-documenting personal knowledge + automation system — Obsidian as canonical store, Claude Code + local LLMs as the activation layer, with drift tripwires as guardrails.

  6. rpg-character-carousel rpg-character-carousel Public

    A LinkedIn 'character sheet' carousel built with a two-stage generative-AI pipeline — Gemini 3 Pro Image for pixel-art items, Claude Design for slide composition.