About
My background is in automotive navigation map data operations, where I have worked with road networks, search and routing quality, and pre/post-deployment validation. I focus less on building software for its own sake and more on defining the QA object, review criteria, and repeatable validation steps that make map data operations easier to inspect and explain.
The center of this portfolio is navigation data QA automation. QGIS Kakao Roadview Bridge connects field-style map review with a desktop GIS workflow. Safety Zone Change Monitoring System connects public safety-zone data changes, operational run history, and map/Roadview based review into a practical data operations project. AI coding tools may support the work, but the main value is domain knowledge, QA workflow design, and verifiable deliverables.
Navigation Data Operations
Road network maintenance, field/Roadview based review, and map data deployment validation.
Map Data QA
Practical QA for search, route guidance, VOC issue analysis, and data change impact review.
GIS Workflow Design
Workflow automation and spatial data analysis tools using QGIS, PostGIS, and Open APIs.
Projects
These projects separate implemented scope, work in progress, and planned extensions. The goal is not to present unfinished work as finished software, but to show how practical navigation data QA problems are structured, tested, and turned into tools.
QGIS Kakao Roadview Bridge
A QGIS Python plugin that connects the QGIS canvas with Kakao Map, Roadview, Local Search, and Kakao Mobility Route to support navigation map QA workflows.
- QGIS canvas synchronization with Kakao Map and Roadview
- Kakao Local place and address search workflow
- Route review using Kakao Mobility Route results
- Route geometry, guidance points, and search results represented as QGIS layers
- Map QA workflow that combines Roadview review with spatial data inspection
Extend the plugin toward a multi-provider route QA workflow by adding Kakao/Naver route comparison and exportable QA result data.
Safety Zone Change Monitoring System
A monitoring project that collects Korean police safety-zone Open API data, stores raw responses, normalized records, current state, and change events in PostgreSQL/PostGIS, and exposes meaningful changes through notifications and a map-based web dashboard.
- Safety-zone API collection, raw snapshot storage, and normalization pipeline
- Separated Polygon, facility Point, and integrated zone group models
- Change events for new, attribute change, geometry change, point change, deletion, and missing-review cases
- Operational SQL checks and change-event verification procedures
- Telegram/Slack notification flow that sends notifications only when relevant changes exist
- Handling rules for empty API responses and 429 rate-limit operational issues
- pytest coverage for API, diff, normalization, pipeline, and notification logic
- Static dashboard data export using JSON/GeoJSON
- GitHub Pages dashboard deployment: safetyzone.yjkim.dev
- OSM/Kakao map switching and Kakao Roadview review for selected locations
- Current layers separated by child, senior, disabled, and other safety-zone types
- Change-state layers for new, changed, and deleted/review-needed objects with popup details
- Recent changes, current objects, and monitoring run history tabs with search/filter UI
- Failed runs are visible with masked sensitive query parameters in failure messages
- Mobile review flow with collapsible bottom panel and draggable/resizable Roadview panel
- Nationwide rollout docs for SGG codes, chunk files, and baseline loading policy
- Prepared 09:00 KST run flow using GitHub Actions self-hosted runner or Windows scheduler
- Documented read contract and QC rules for the existing standard node-link PostGIS database
- Using nationwide NGII road centerline and 0.5m simplified geometry as a supporting reference
- Designed candidate, excluded, and coverage review views for safety-zone to road-link matching
- v2.3 review buckets separate auto-apply candidates from manual-review candidates conservatively
- QGIS styles and casebooks record elevated roads, underpasses, roundabouts, near-parallel false positives, and missing-link patterns
The dashboard is deployed as an operational review artifact, but nationwide baseline database loading still needs chunk-based retries and stabilization due to API 429 limits. Standard-link matching is still in candidate generation and QGIS review criteria refinement, not automatic confirmed matching.
Multi-Provider Navigation Route QA System
A planned extension of the QGIS Kakao plugin to compare Kakao and Naver route results and analyze route quality and guidance differences.
Planned Scope- Compare Kakao / Naver route geometry
- Analyze distance, duration, and turn-by-turn guidance differences
- Review route divergence, missing guidance, and abnormal detour candidates
- Use the QGIS plugin as the desktop analysis tool
- Consider a future web dashboard for exported route comparison results
Blog
This space is intended for technical notes on designing and validating GIS, QGIS, and PostGIS-based QA tools grounded in automotive navigation data operations.
- Why I built a QGIS Roadview-based map QA workflow
- Safety-zone Open API change monitoring MVP and operational validation criteria
- Nationwide baseline loading and interpretation of new/changed/deleted events
- Safety-zone change review dashboard using OSM/Kakao maps and Roadview
- Standard-link matching v2.3 review buckets and QGIS visual QA cases
- Why Kakao / Naver route comparison matters in navigation QA
Contact
If you would like to discuss map data QA, road network validation, mobility data operations, or GIS-based workflow tooling, please contact me below.
Portfolio: yjkim.dev
Email: contact@yjkim.dev
LinkedIn: www.linkedin.com/in/youngj-kim
GitHub: github.com/youngj-kim