AI Backend Engineer - Data & Integration

## About the role At Digital Iron, we're building the intelligent infrastructure that powers predictive maintenance and parts procurement automation across the heavy equipment ecosystem. We work with customers to transform how industrial equipment is maintained. We're looking for an AI Infrastructure Engineer who combines deep technical expertise in distributed systems with strategic thinking about integration architecture. You'll need exceptionally high standards for data accuracy, first-principles problem solving, and an obsession with building systems that scale across diverse partnership models. As our first dedicated infrastructure engineer, you'll work on problems at the intersection of knowledge graphs, real-time IoT data, and enterprise integration—building infrastructure that thousands of businesses will depend on. ## What You'll Own - Design the framework that supports multiple partnership and customer models: Deep Embedded (white-label components), Best-of-Breed SaaS (standalone platform with APIs), Data Layer Only (predictions via API) - Evaluate architectural tradeoffs across complexity, risk, scalability, time-to-market, and value capture for each pattern - Make build vs. buy decisions: direct API integrations vs. iPaaS middleware vs. embedded agents - Define authentication strategies across OAuth 2.0, certificate-based auth, and federated identity for different customer security models - Create deployment patterns that work across on-premise, cloud, and hybrid environments - Design for portfolio risk ## What You'll Do - Design Integration Architecture: Build bi-directional integrations with customer ERP systems and telematics platforms. Architect event-driven systems that turn predictive alerts into automated workflows. Implement multiple integration patterns (Direct API, middleware/iPaaS, embedded agents, webhooks) to support different partnership and customer models. - Build Knowledge Graph Systems: Transform flat parts catalogs into semantic networks using AWS Neptune. Design ontologies that capture ACES (fitment) and PIES (attributes) standards for heavy equipment. Build ingestion pipelines that parse customer data and extract compatibility relationships. Implement graph traversal algorithms for multi-hop reasoning ("find compatible substitute parts in stock"). - Develop Agentic Workflows: Create AI agent orchestration using tools like Amazon Bedrock that breaks complex requests into multi-step workflows. Build tool functions agents invoke: graph queries, customer API calls, inventory checks, order placement. Implement GraphRAG systems that ground LLM responses in structured graph data to prevent hallucination on critical fitment recommendations. ## What We're Looking For ### Graph & Semantic Systems - Experience with graph databases (Neptune, Neo4j) and ontology design - Ability to model complex domain relationships as graph structures - Understanding of semantic query languages (Gremlin, SPARQL) and entity resolution ### Backend Engineering & Data Systems