Junho Lee (이준호)

Data & AI Platform Architect | PM

I have designed and led large-scale DW cloud migrations and next-generation data platform builds on the front lines of DW/BI delivery. My career started as a Web/ERP developer, then progressed through DW/BI engineer, Technical Lead, and Consulting Division Head — and now I’m building an ontology-based AI data platform.

Career Summary

The first half of my career was focused on enterprise DW/BI. I led large-scale DW cloud migrations as a Tech Leader and operated next-generation information system projects as a multi-vendor PMO. I’ve worked across a wide range of industry domains including retail, telecommunications, manufacturing, and construction.

I also have experience building a consulting organization from the ground up — growing a small team to 20+ people and scaling revenue several times over. Hiring, training, managing a technical organization, Presales, C-level seminars — I’m not just someone who codes.

More recently, I’ve been working at the intersection of data and AI. While building an LLM-based BI Agent, I encountered the real-world limitations of NL2SQL firsthand, and through that process became convinced of the need for an ontology-driven approach. I am now designing and building DataNexus, an integrated data agent platform.

DataNexus

“Everyone is an Analyst.”

A platform designed to solve the structural problems of enterprise data analytics. An AI agent that lets anyone explore and analyze internal data through natural language — easy to say, but in practice, table names look like T_CUST_MST, full of abbreviations, and a single term like “net revenue” carries different calculation logic across departments. LLMs cannot understand business context from DDL alone.

DataNexus tackles this problem by combining an ontology-based NL2SQL engine, GraphRAG, and a Data Catalog. Built on an open-source composite architecture, it provides a single interface for handling both unstructured documents and structured databases.

This blog documents the process of building DataNexus — architecture decisions, reasons behind technology choices, and the struggles and solutions along the way, recorded as-is.

Technical Areas

  • AI/ML — Ontology LLM RAG, NL2SQL, Langchain, MCP, multi-agent system design
  • DW/Data Platform — Azure Synapse, BigQuery, Redshift, PostgreSQL, Oracle, Yellowbrick, Palantir Foundry
  • BI — Power BI, Tableau, MicroStrategy, Qlik Sense, Looker, Superset
  • ETL/ELT — ADF, SAP Data Services, IBM DataStage, Informatica, Databricks, SSIS
  • Cloud — Azure (Synapse, ADF, ML), AWS (Redshift, S3, Glue), GCP (BigQuery, Gemini)
  • Graph/Catalog — DataHub, Neo4j (DozerDB), ApeRAG

Contact