Industrial

Agentic NL 3D Model Search: LangGraph Pipeline

Agentic Compositional Multimodal Natural Language 3D Model Search

Building an agentic multimodal search system (CoI-Fit) for 3D product retrieval at Bucketplace, combining LLM reasoning with BM25+KNN hybrid search across visual, textual, and spatial signals via LangGraph and A2A protocol.

Segment Selection and 3D Layout Visualization

Scene-to-Products Retrieval for Digital Twin (Image-to-3D)

Built an end-to-end scene-to-products retrieval pipeline for Digital Twin at Bucketplace, achieving 13x recall improvement, 6,000x cost reduction in 3D collaboration, and scaling the catalog from 30K to 400K+ products.

OHouseAI GenAI Workflow Systematization

Led systematization of OHouseAI’s GenAI interior design service at Bucketplace, achieving +253% net satisfaction, 2x latency reduction, and outperforming GPT-IMAGE-1.5 and Gemini Nano. Reached Korea #8 in Graphics/Design.

Personalized Multimodal Retrieval Pipeline

Personalized Multimodal Retrieval with VLM Distillation

Developed personalized multimodal retrieval combining SASRec sequential recommendation with Jina-CLIP-V2 vision-language representations, using knowledge distillation for efficient production serving at Bucketplace.

A/B Test Results for Search OKR Experiments

Search OKR: E-commerce Deals / Price 2.0 Ranking Optimization

Delivered measurable search ranking OKR impact at Bucketplace through 4 production A/B experiments (3 winners shipped), achieving +11.17% buyer conversion on category pages and +7.35% deals exposure on store search.

MOHPER Framework Architecture (Figure 1, CIKM Paper)

Multi-Objective Model Hyperparameter Optimization for E-commerce Search

Built a multi-objective hyperparameter optimization framework for e-commerce search at Bucketplace, launching tuned parameters 5x+ on SERP and 3x+ on CPLP for CTR/CVR gains. Accepted to CIKM'25 Applied Research Track.

Global Content Search Two-Stage Ranking Pipeline

Global Content Search Ranking System

Built a two-stage content ranking system at Bucketplace covering 3 content types, with ElasticSearch function-score reranking, time decay, global language analysis, and Redis-cached serving.

CLIP Multimodal Retrieval: Before vs After — zero-result queries now return relevant products via semantic matching

CLIP-based E-Commerce Multimodal Search and Classification

Shipped CLIP-based multimodal retrieval at Bucketplace, achieving +16.39% query CTCVR, 2x inference speed via quantization, and 92.23% Top-10 accuracy across 2,000+ product categories.

Shared ML Feature Mart & Experiment Infrastructure

Design & Development of a Shared Team ML Mart

Designed a shared ML feature mart for Bucketplace Search Team with client-side log validation, lag anomaly detection, and automated A/B test analysis pipeline using PySpark and HiveDB.

Figure 3: End-to-end speech scoring model architecture from InterSpeech paper

Prompt-Aware Speech Scoring System for Second Language Learners

Developed a prompt-aware automatic speech scoring system at RIIID that solves the cold-start item problem using BERT/CLIP prompt embeddings. Published at InterSpeech 2023 and deployed to SANTA Say TOEIC Speaking app.