Skip to content
View LearningnRunning's full-sized avatar
πŸ€“
πŸ€“

Organizations

@A1Platform @Catch-V @lunch-corp

Block or report LearningnRunning

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
LearningnRunning/README.md

Hi, I'm Rocky πŸ‘‹

ML/AI engineer working on AI agents and ML systems in the advertising domain.

I started programming in 2022, later than most, and have kept building side projects alongside full-time work since then.

  • πŸ”­ Currently working on ad copy compliance checking (detecting regulatory violations) and ad platform backend/batch systems at an advertising agency, 4 years in.
  • 🌱 Outside of work, I build side projects across different domains β€” restaurant recommendation systems, a financial data MCP server, a task management app, an AI tutoring simulator.
  • ⚑ I favor a build β†’ validate β†’ ship cycle: turning ideas into working prototypes, then testing them in production.
  • πŸ’¬ I like owning the full stack of a problem β€” from ML modeling to backend to deployment.

Work

Ad Copy Compliance Checker Rebuilt a legacy binary classifier into a RAG + multi-model ensemble system with a continuous retraining loop. Improved violation detection recall and automated both the reasoning behind each flag and suggested rewrites.

Ad Platform Backend/Batch System Owned the backend of a production-scale ad platform for over a year β€” reach estimation, large-scale batch upload validation, and integrations with external systems.


Side Projects

Jjeopjjeop LAB / TasteRoute β€” Review-based Restaurant Discovery

A restaurant recommendation service built on a semantic pipeline that normalizes public review data, with a Bayesian-adjusted ranking model accounting for review reliability, recency, and engagement. Currently in production. Also designing a Semantic ID–based generative recommendation architecture (TIGER-style) as a separate track.

trade-bridge β€” US Stock Research MCP Server

An MCP server exposing 57 tools for LLMs to query and combine macro indicators, financial statements, sector momentum, SEC filings, DCF models, and options data. Deployed and in active use on Oracle Cloud's free tier.

Dear J β€” A Todo App Connecting Projects to Daily Execution

Built around the idea that choosing, doing, and reflecting on "what to do today" across multiple projects is itself the core product experience. Designed, built, and deployed end-to-end with React and Supabase.

Nursing Handoff SBAR Simulator β€” AI Tutoring Simulator

Built with a PhD candidate in nursing education. Separates evaluation from response generation, using a Socratic approach that withholds answers so students identify gaps in their own clinical reasoning. Currently piloting with the nursing department at Mokpo National University.


Contact

max_sungrok@naver.com


Popular repositories Loading

  1. NaverTrans NaverTrans Public

    넀이버 λ²ˆμ—­κΈ°λ₯Ό ν™œμš©ν•œ 파이썬 νŒ¨ν‚€μ§€μž…λ‹ˆλ‹€.

    Python 6

  2. NBus NBus Public

    Python 4 2

  3. What2Eat-legacy What2Eat-legacy Public

    Python 3 2

  4. LearningnRunning LearningnRunning Public

    This is My Profile

    1

  5. Midnight_Diner Midnight_Diner Public

    Forked from jjy0328/Midnight_Diner

    심야식당 ν”„λ‘œμ νŠΈ ν˜‘μ—… λ ˆν¬μ§€ν† λ¦¬μž…λ‹ˆλ‹€.

    Jupyter Notebook 1

  6. Algorithm Algorithm Public

    Algorithm every week.

    Python