This role is with one of Dex’s trusted partner companies. We work closely with their teams to truly understand their culture, goals, and what they’re looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process.
The role
The physical economy—from automotive spare-parts distributors to heavy equipment and salvage operators—runs on decades of fragmented legacy ERPs, on-prem databases, and unstructured file dumps. Fresh off a major round from top-tier VCs, this company was founded by AI researchers to build the unified AI operating system that turns this messy data into automated pricing, inventory, and distribution intelligence.
You will partner directly with the founders to build the platform from the ground up. This is not an operational maintenance role or a narrow feature ticket queue; you will own end-to-end data ingestion, workflow orchestration, and backend service architecture, while defining the company standards for data quality, schema evolution, and deployment pipelines.
The work
- Architect and operate resilient ingestion pipelines that ingest, clean, and transform messy customer data across legacy ERPs, on-prem databases, and flat files.
- Design and maintain observable workflow orchestration systems using DAGs to keep critical data pipelines running reliably.
- Build robust, scalable Python APIs and microservices that power AI-driven insights and customer-facing interfaces.
- Set engineering standards across the stack, including data quality gates, schema evolution strategies, CI/CD automation, and observability.
What You Bring
- Substantial production experience designing, building, and operating backend systems and APIs in Python, with a track record in early-stage or high-velocity engineering teams.
- Hands-on ownership of end-to-end data pipelines and workflow orchestration engines such as Airflow or custom DAG-based systems.
- Deep experience tackling messy integrations across legacy protocols, on-prem databases, SFTP feeds, and unstructured file dumps.
- High agency and comfort operating with early-stage startup intensity, shaping architectural foundations without relying on predefined playbooks.
#J-18808-Ljbffr…