Summary: Seeking a Senior Backend Engineer to build and own core services for an AI-native platform, focusing on APIs, data modeling, event-driven workflows, and reliability. Highlights: 1. Build and own core backend services for an AI-native platform 2. Design clean APIs and data models across multiple databases 3. Integrate LLM/embedding services with deterministic rules PLEASE APPLY AT: https://app.planetarytalent.com/apply?role\=d59fb936\-59d7\-45a2\-849e\-041c47f2ba64 About the Platform AI\-native system for turning regulatory evidence into market\-ready packets. Tracks proofs across product × site × market, orchestrates supplier campaigns, validates document lineage/expirations, encodes rules (RoHS, REACH, TSCA, PFAS, CE/UL/CSA, EPD/EN 15804, DoP/CPR), and ships one\-click outputs to customers, auditors, and borders. Automatic. Scalable. Real\-time. The Role Senior Backend Engineer building and owning core services: evidence graph, supplier pipelines, rules/validations, and packet factory. Heavy on APIs, data modeling, event\-driven workflows, and reliability. Work tight with product and AI to ship fast, measure impact, and keep SLAs sharp. Modern stack, real ownership, high leverage. What You’ll Do * Build product end\-to\-end. Design, implement, and ship backend services (Node.js/TypeScript, \- Python) that ingest documents, validate evidence, and generate market\-ready packets. * Own APIs and data models. Design clean REST/GraphQL APIs; model data across Postgres/Aurora (relational), DynamoDB (document), and S3; enforce provenance and audit trails. * AI \+ guardrails. Integrate LLM/embedding services with deterministic rules; wire RAG pipelines, citations, confidence thresholds, and human\-in\-the\-loop review paths. * Rules and validations. Encode checks for RoHS/REACH/TSCA/PFAS, UL/CE/IEC/CSA, NSF 61/372, DoP/CPR, EPD/EN 15804; implement versioning and diffs. * Reliability at scale. Use eventing/queues (SNS/SQS/Kinesis), serverless (Lambda) and containerized services to process large doc volumes with strong SLAs. * Security and compliance. Build with least privilege, secrets hygiene, and logging to support SOC 2/GDPR; contribute to threat modeling and privacy reviews. * DevEx and quality. Add tests (unit/integration/e2e), CI/CD (GitHub Actions), feature flags, and deep observability (DataDog/OpenTelemetry) to keep the system fast and debuggable. * Integrations. Ship connectors to ERP/PLM (SAP, Oracle, Teamcenter, Windchill), identity (SSO/SAML/OIDC), and content stores; later, push packets to routing tools. * Own outcomes. Partner with PM/design to scope, instrument, and iterate measuring minutes\-to\-packet, extraction precision, and time\-to\-first\-value. Our Stack (today) * Frontend: React, TypeScript, Next.js, Tailwind, Vite * Backend: Node.js (TypeScript), Python (for AI/ETL), REST/GraphQL, gRPC (select services) * AI/ML: embeddings \+ LLM orchestration (LangChain/LangGraph\-style patterns), vector store, OCR/layout parsing * Data \& Infra: Postgres/Aurora, DynamoDB, S3, Step Functions/Lambda, SNS/SQS, Terraform, DataDog, OpenTelemetry, CloudFront * DevOps: GitHub Actions, IaC, feature flags, preview envs What Success Looks Like (first 90–180 days) 90 days: * Ship a customer\-visible workflow end\-to\-end (UI \+ API \+ data) with tests and dashboards. * Reduce a packet flow from hours to \<10 minutes wall\-clock in production. * Land one integration (e.g., supplier intake or ERP/PLM export) with robust retries/observability. 180 days: * Stand up a reusable evidence graph module (provenance, versioning, expiry watch) used by multiple features. * Improve extraction quality with guardrails (measured precision/recall on key fields); cut rework by 25–40 percent. * Author or own a service with 99\.9 percent plus monthly availability and SLO dashboards. What You’ll Bring Must\-Have * 6 plus years building production SaaS with modern JS/TS plus a typed backend (Node.js, Go, or similar) and practical Python for data/AI tasks. * API and data design chops (REST/GraphQL, SQL/NoSQL), event\-driven patterns, and cloud experience (AWS preferred). * Experience with at least one of: document processing/OCR, LLM/RAG, or complex workflow engines with reliability/latency concerns. * Track record shipping measurable improvements (perf, reliability, product adoption) in an agile environment. Nice\-to\-Have * LangChain/LangGraph patterns, vector databases, prompt/guardrail tooling. * PDF/layout parsing, table extraction, entity resolution. * ERP/PLM or compliance domain exposure (RoHS, REACH, TSCA, PFAS, CE/UL/CSA, EPD/DoP). * Terraform/IaC, DataDog/Otel, Temporal/Step Functions, Auth (SAML/OIDC), secure file handling. PLEASE APPLY AT: https://app.planetarytalent.com/apply?role\=d59fb936\-59d7\-45a2\-849e\-041c47f2ba64 Job Type: Full\-time Application Question(s): * What is the most complex software engineering problem/project that you have worked on? Experience: * software engineering: 5 years (Required) Work Location: Remote