Let's Talk

Feel free to reach out. I'll get back to you as soon as possible.

Jaehwan Park

Profile Image
Professional Profile

Professional in the Cloud Services industry focused on data center stability and core infrastructure reliability. Combining self-taught programming foundations with systematic professional training to deliver technical excellence and effective team collaboration.

🛠 Skills & Experience
Languages & Tools
  • Languages: C, Python (Pandas, NumPy, Scikit-learn)
  • AI/ML Frameworks: TensorFlow
  • Security & Analysis: Burp Suite
  • Adaptability: Proven ability to quickly master and implement diverse technical stacks.
Key Achievement: AI Project
  • 2022 AI Online Challenge: Developed a beef quality classification model.
  • Impact: Successfully improved model accuracy from 50% to 89% through optimization and data analysis.
Education
  • K-Digital Training (National Competency Standards): Completed as an Outstanding Trainee.
  • Collaboration: Led and participated in various team projects, fostering strong communication and cooperative problem-solving skills.
🔭 Focus & Interests
  • Infrastructure & Operations: Committed to the stable management of data center servers and high-availability network operations.
  • Continuous Growth: Aiming for proactive task execution by rapidly acquiring new technologies and internalizing operational manuals.
  • Responsibility & Integrity: A reliable team member built on meticulous time management and professional sincerity.
🚀 Ongoing Toy Projects
1. RAG (Retrieval-Augmented Generation) System (In Progress)

Developing a specialized search and Q&A system for gaming internal resources (dialogues, items, regional lore).

  • Technical Detail: Vectorizing unstructured game data into a Vector Database to enable LLMs to provide context-aware, accurate responses.
  • 1-1. TestRAG (Sub-project): A diagnostic tool for measuring retrieval performance and visualizing vectorized data using Three.js.
2. RIDI: Local LLM-based Assistant (In Progress)

A personalized AI assistant inspired by Neuro-sama and Sovits_GPT, focused on local execution (Model size ≤ 8B).

  • Fine-tuning English models for enhanced Korean proficiency.
  • Voice interaction and response generation based on long-term memory and system prompts.
  • Integration with local system resources for PC control and manipulation.
3. MoneyView (Planning AI Agent Integration)

A comprehensive information aggregation tool for tracking rapidly changing trends in Economics, Society, Politics, and Diplomacy.

Future Roadmap: Integrating AI Agents to autonomously categorize and summarize global intelligence.

💡 Vision & Technical Philosophy

As the era of AI Agents (such as OpenClaw, Qwen 3.5, GLM 4.7, Claude Coworker, and Gemini AntiGravity) emerges, I believe that mastering AI integration is no longer optional but essential. My current focus is on:

  • Information Architecture: Beyond simple search engines, I prioritize the systematic management of the underlying data that powers AI.
  • Agile Development with AI: Actively utilizing Code Assistants to accelerate development speed while maintaining a deep commitment to understanding the core principles required to build truly innovative systems.
  • Lifelong Learning: Recognizing that creating "the new" requires profound knowledge, I am dedicated to bridging the gap between existing tools and groundbreaking implementation through continuous study and experimentation.