KoongLog는 공동주택의 층간소음 데이터를 수집·분석하고, 관리자와 주민이 갈등을 조기에 파악하고 중재할 수 있도록 돕는 AI/IoT 플랫폼입니다.
This repository presents my backend contribution to the KoongLog team repository as a concise portfolio case study.
flowchart LR
A[IoT sensor or simulator] --> B[FastAPI ingestion]
B --> C[Noise classification service]
B --> D[(SQLite database)]
C --> D
D --> E[Dashboard analytics]
D --> F[Hotspot and risk monitoring]
D --> G[AI mediation workflow]
E --> H[Administrator web]
F --> H
G --> H
G --> I[Resident app]
The service connects sensor ingestion, noise-event analysis, household data, dashboard analytics, and an AI-assisted mediation workflow in one backend.
- Expanded the SQLAlchemy data model for raw sensor readings, noise events, household information, and mediation data.
- Built dashboard APIs for monitored households, urgent cases, daily events, completed actions, and hourly statistics.
- Implemented noise-distribution, hotspot, household-summary, resident-profile, and 24-hour statistics endpoints.
- Added mediation-request, approval, scheduling, notice, report, and PDF-related backend workflows.
- Integrated Arduino-oriented ingestion changes and improved timestamp and Korean Standard Time handling.
- Resolved merge conflicts and aligned API response fields with frontend needs.
These contributions are traceable in 75 public commits authored under
@hozziii.
| Area | Examples |
|---|---|
| Sensor and noise | Sensor ingestion, recent readings, distributions, hotspots |
| Dashboard | Household monitoring, urgent cases, hourly and daily statistics |
| Mediation | Requests, detail views, status updates, approval and delivery |
| Residents | Home summary, profile, quiet time, withdrawal, 24-hour statistics |
| Operations | Notices, AI templates, reports, PDF export, schedule management |
| Area | Tools |
|---|---|
| Backend | Python 3.10+, FastAPI, Pydantic |
| Data | SQLite, SQLAlchemy ORM |
| AI integration | Noise-analysis service, LLM-assisted mediation messages |
| Infrastructure | Railway, Uvicorn, Arduino sensor/simulator input |
This personal repository is a documentation-first portfolio. The maintained implementation, complete history, setup instructions, and API documentation are available in the original backend repository. No team code or credentials are duplicated here.
- Translating product requirements into data models and versioned REST APIs
- Designing analytics endpoints over time-series and household-level data
- Integrating IoT input, AI services, and operational workflows
- Collaborating on a fast-moving backend with pull requests and schema changes