ANTHROPIC SEOUL

Application Packet

Doeon Kim — Solutions Architect / Data Platform Engineer

Anthropic Alignment

stage-pilot — Tool-calling reliability runtime (npm: @ai-sdk-tool/parser, 1,720 tests) with first-class Anthropic/Claude model support and streaming structured outputs.
enterprise-llm-adoption-kit — LLM governance framework with RBAC, PII redaction, eval pipelines — embodying Anthropic's commitment to responsible AI deployment.
tool-call-finetune-lab — QLoRA fine-tuning and BFCL benchmark evaluation, advancing tool-use reliability research.
AegisOps — Multimodal incident analysis with safety-conscious design, live demo deployment.

Technical Depth

Tool use: Published reliability runtime with streaming parsers, structured output validation, 1,720 tests
Safety & governance: PII redaction, RBAC, eval frameworks, audit logging for responsible AI
Fine-tuning: QLoRA, PEFT, benchmark-driven evaluation of tool-calling capabilities
Agent systems: LangGraph multi-agent orchestration, governed analytics, NL-to-SQL
Full-stack: TypeScript/Python, React, FastAPI, Spark — production-ready at every layer

Why Anthropic Seoul

I build AI systems with reliability and safety as first-class concerns. My tool-calling runtime, governance frameworks, and eval pipelines reflect the same principles that drive Anthropic's mission to develop safe, beneficial AI.

Credentials

Microsoft AI-900
SnowPro Associate & Databricks Platform Architect
Palantir Foundry Data Engineer + Foundations
Datadog Observability Fundamentals
Trilingual: Korean (native) / Japanese / English — Seoul based

stage-pilot LLM Adoption Kit