Doeon Kim — Solutions Architect / Data Platform Engineer
AWS & GenAI Alignment
stage-pilot — Tool-calling reliability runtime (npm: @ai-sdk-tool/parser, 1,720 tests) designed for production AI workloads deployable on AWS Lambda and ECS.
enterprise-llm-adoption-kit — LLM governance framework with RBAC, evals, PII redaction, and FastAPI — architected for Bedrock and SageMaker integration.
lakehouse-contract-lab — Spark + Delta Lake medallion pipeline with data contracts, designed for EMR and Glue workloads.
Technical Depth
LLM fine-tuning: QLoRA on Qwen2.5-7B with BFCL benchmark evaluation (SageMaker-ready)
Multi-agent orchestration: LangGraph agents with tool-calling, NL-to-SQL, governed analytics
Data platform: Spark, Delta Lake, medallion architecture, Snowflake/Databricks integration
Production AI: Streaming parsers, structured output validation, reliability engineering
Observability: Datadog certified — metrics, traces, logs for AI workloads
Why AWS GenAI
I bring production-tested LLM tooling and data platform expertise that maps directly to helping AWS customers adopt Bedrock, SageMaker, and generative AI services at scale.