SNOWFLAKE KOREA

Application Packet

Doeon Kim — Solutions Architect / Data Platform Engineer

Snowflake Alignment

DistrictPilot AI — 100% Snowflake Native budget-allocation decision agent built for the Snowflake Korea Hackathon 2026 Tech Track. Uses Cortex ML Forecast, Cortex Analyst (Semantic View), and Streamlit in Snowflake.
Nexus-Hive — Multi-agent NL-to-SQL copilot with governed analytics and multi-warehouse support including Snowflake.
Lakehouse Contract Lab — Spark + Delta Lake medallion pipeline with Snowflake warehouse export and data contracts.

Technical Depth

Snowpark Python, Cortex ML Functions (FORECAST, ANOMALY_DETECTION)
Semantic Views with VQR for Cortex Analyst
Streamlit in Snowflake (owner's rights, multi-tab SPA pattern)
Data engineering: medallion architecture, staged ingestion, SQL-based feature marts
ML/AI: PyTorch fine-tuning (QLoRA), LangGraph agents, tool-calling reliability

Why Snowflake

I build end-to-end AI solutions on a single governed platform. DistrictPilot AI proves that forecast, analysis, simulation, and explainability can all run inside Snowflake without external dependencies.

Credentials

SnowPro Associate
Microsoft AI-900
Microsoft AI School Graduate
Trilingual: Korean (native) / Japanese / English

View DistrictPilot AI on GitHub