Kangnam Kim

May 2026 - Present

Rye&Catchers

AI Research Engineer — AI Research Team

Security Anomaly Detection Systems

• [Security] Train and validate deep-learning anomaly-detection systems for first-stage quarantine of malicious behavior and PE files
• [Anomaly Detection] Apply LSTM likelihood-based scoring (NLL) to CERT user-behavior data and VAE reconstruction error with uncertainty quantification to BODMAS PE-file data

2024.04 ~ 2025.06

Creverse (Chungdahm Learning)

AI Engineer — Nesto CV Division, AI Biz Team

AI Auto-Assessment System

• [MLOps] Designed, deployed, and operated an Azure ML Endpoint and Blob Storage inference pipeline used for presentation assessment by approximately 30K students monthly
• [LLM Agent] Built LangGraph student-and-evaluator agents to generate and review grade- and proficiency-conditioned synthetic English writing for international assessment exploration
• [Audio] Fine-tuned NVIDIA Parakeet STT on non-native children's speech data, reducing WER from 9.6% to 4.4%
• [Vision] Built MediaPipe-based gaze and posture assessment with 98.32% validation accuracy and 98.77% F1, configured for practical CPU inference
• [Vision] Used a quantized image-generation model during development to reduce GPU cost; added AWS image-safety detection and initial human review for a children's service
• [Search] Combined BM25 and vector retrieval; internal log-based evaluation showed up to ~20% fewer LLM calls at the same quality threshold

2021.06 ~ 2022.04

Coconut Silo

AI Research Engineer — CTO Office

• [Recommender System] Embedded product descriptions using BERT and implemented a content-based filtering (CBF) recommendation system using inner product similarity
• [Vision] Fine-tuned YOLOv5 for ship and obstacle detection in maritime radar imagery
• [Anomaly Detection] Designed and implemented a vision-based anomaly detection system for real-time analysis of metal 3D printer process footage