System-level Software Engineer specializing in Database Engine Kernels & Distributed Coordination
- System-level software engineer specializing in database engine kernels and distributed coordination.
- Expertise in extending PostgreSQL with C extensions, hooks (planner, object access), Vector/LLM integration, and Oracle-compatible features.
- Experience architecting distributed lifecycle engines with a focus on transactional state consistency and reliability.
- Proficient in C, C++, Python, and Linux; solid foundation in algorithms, probability, and high-performance distributed system design.
| Area | Details |
|---|---|
| Languages | C, C++ (Modern), Python, SQL, Java, Bash |
| DB Internals | PostgreSQL/Oracle engine core, query compiler, optimizer (CBO), vector search |
| PostgreSQL | C extensions, planner hook, object access hook, RLS, parser/planner integration |
| Distributed | Consensus (etcd/Raft), high availability, cluster orchestration (IaC) |
| Core | Probability, high-performance algorithms |
| Tools | Linux (system-level), Docker, Git, AWS/Azure (SDK/API integration) |
1) OpenSQL Team (Nov 2024 โ Present)
PostgreSQL-based managed RDBMS with Oracle compatibility and vector DB features.
-
Row-Level Security (DBMS_RLS)
Designed and implemented Oracle-compatible RLS as a PostgreSQL C extension, using planner_hook and object_access_hook for planner integration and runtime policy enforcement. Delivered transient-viewโstyle predicate injection and definer-rights policy execution aligned with Oracle VPD semantics. -
Oracle compatibility
Implemented Oracle-compatible functions/packages and conducted performance benchmarking; achieved measurable improvements including 50%+ gain for the median function through algorithm redesign. -
Vector DB & RAG
Integrated pgvector, developed vector embedding features, and implemented LLM integration for DB-based RAG solutions. -
High availability
Designed PostgreSQL HA environments using Patroni and etcd for enterprise workloads. -
Cluster orchestration & scaling
Architected a Multi-CSP IaC (ARM/CloudFormation) scaling engine using etcd watch for real-time state sync and Command Patternโbased rollback for transactional reliability in distributed node lifecycle. -
Debug / monitoring
Built internal debug tooling to print PostgreSQL query trees during development.
2) SuperTibero Team (Oct 2023 โ Oct 2024)
Enterprise RDBMS with distributed storage; focused on SQL compiler modules.
-
SQL compiler
Built parser, query transformer, and cost-based optimizer. -
Query optimization
Implemented index range scan, predicate pushdown, sort-skip optimization, and execution plan cache. -
Privilege & authentication
Designed and implemented privilege check and authentication for secure SQL execution. -
Performance
Delivered query transformation features that improved index scan opportunities and reduced redundant operations.
- B.S. in Physics โ Yonsei University, Seoul
- Focus: Statistical Mechanics, Quantum Mechanics
-
Algorithms
-
Certifications
- Engineer Information Processing (HRDK)
- SQL Developer (SQLD, Kdata)