Knowledge Data Engineer

TEKsystems•Charlotte, NC

11d•$65 - $72•Hybrid

About The Position

Support the build out of the process management semantic layer. The role focuses on logically connecting business terminology to Oracle database structures, using PDFs, Oracle metadata via Toad, and Python to define business definitions, synonyms, and term-to-data mappings. Over time, the role may extend to documenting basic rules and relationships. The Knowledge / Data Engineer is responsible for building the semantic bridge between process management business language and Oracle data structures. This role converts business terms, definitions, synonyms, source metadata, and documentation into structured mappings that help GenAI interpret user questions and generate accurate database queries.

Requirements

Data
Data architecture
Data modeling

Responsibilities

Review process management PDFs, business documents, data dictionaries, and metadata extracts to identify business terms and domain concepts.
Define clear business definitions for terms used by process management users.
Identify synonyms, abbreviations, alternate names, and commonly used phrases for each business term.
Distinguish between similar or overloaded terms that may map to different database structures depending on business context.
Maintain terminology consistency across the semantic layer and knowledge base repository.
Use Toad and Oracle metadata extracts to understand available schemas, tables, columns, keys, and relationships.
Analyze physical data structures and determine which tables and columns support each business term.
Identify candidate joins, filters, dimensions, measures, and identifiers needed to answer common process management related questions.
Work with the Data Architect to validate table relationships, materialized views, and performance-optimized access paths.
Create mappings from business terms to Oracle tables, columns, views, and materialized views.
Define mapping metadata such as: Business term, Business definition, Synonyms, Source schema, Source table or view, Source column, Data type, Join path, Filters or conditions, Valid values, Example user phrases, Confidence level, Source documentation reference.
Ensure mappings are precise enough to support natural-language-to-SQL generation.
Identify gaps where business terms lack clear physical data mappings.