AI Briefing
Naver Place news
The latest AI and developer news from Naver Place, with short summaries.
Visit website →Building a Backoffice AI Agent: A RAG+MCP-based Knowledge Search System Tailored for PlaceAI
PlaceAI automated backoffice knowledge search using RAG+MCP and hybrid search.
Optimizing Trino Resources on YARN
By tuning YARN's RESERVED Resource and Trino settings, available memory was increased by about 40%.
Naver Map Navigation, Smooth Roads Only: A Narrow Road Detection Model
Using digital topographic maps and a Street View SegFormer, narrow roads were identified more accurately.
Avoiding Airflow Task Failure Alert Email Bombs
By bundling per-Task failure emails in Airflow into a single message, alert noise was reduced.
Containerizing the Airflow Development Environment with Docker Compose
Docker Compose was used to align the local Airflow environment more closely with the deployment environment, enabling debugging in PyCharm as well.
Improving the Place Review Tagging System to Achieve a 67% Update Compression Rate
By separating tagging logic from storage and adding a write buffer, updates were reduced by 67%.
Introducing NAVER Place's AI Development Team
NAVER GLACE is upgrading its place and review services with 15+ AI models and PlaceLM.
Experience Migrating an ML GPU Model Server to a CPU Server While Maintaining Performance
Boosted CPU serving performance 4x through torchserve thread configuration and ipex pinning.
Streamlining Development with GitHub Actions
Used GitHub Actions and Marketplace actions to automate PR approvals, notifications, and syncing.
MinIO Adoption Journey - Understanding HA and Establishing a DR Strategy
HA was verified with 4-node distributed MinIO, and archived data was operated with Backup & Recovery.