This is a remote position.

US – Data Engineer (Pipelines & Structured Markup)

Title: Data Engineer – Pipelines & Structured Markup

Location: US (Part Time, Remote or Hybrid)

Company: Vulcury LLC

Role Overview

Vulcury is building a manufacturing intelligence infrastructure that converts raw interactions — interviews, transcripts, CAD uploads, commercial discussions — into structured, queriable data objects.

We are seeking a Data Engineer to design and maintain ingestion pipelines and structured transformation workflows that power our internal semantic “truth layer.”

This is not a reporting role.

This is a semantic infrastructure role.

Responsibilities

Build and maintain ingestion pipelines (Python-based ETL/ELT)
Design structured transformation workflows using dbt, SQLMesh, or equivalent
Convert unstructured transcripts and documents into normalized database records
Maintain PostgreSQL architecture (structured tables, JSONB, indexing strategy)
Develop attribute extraction frameworks for technical, commercial, and risk signals
Ensure data quality, consistency, and lineage from raw interaction to structured output
Collaborate with AI/ML engineers to ensure clean model inputs

Requirements

Required Skills

Strong Python (data pipelines, orchestration)
Advanced SQL (PostgreSQL preferred)
Experience with ETL/ELT frameworks (dbt, Airflow, SQLMesh, etc.)
Experience handling semi-structured data (JSON, transcripts, document parsing)
Strong schema design and normalization skills
Familiarity with cloud storage systems (S3 or equivalent)

Nice to Have

Experience building semantic layers or knowledge graphs
Experience working with manufacturing or technical data
Familiarity with vector databases

Benefits

What Success Looks Like

Raw interviews automatically convert into structured records
Attribute confidence scoring flows downstream cleanly
Data lineage is fully traceable
Query performance remains stable as data volume scales

Vulcury Venture Studio and Advisors

New Jobs

US – Data Engineer (Pipelines & Structured Markup)