Data Platform Engineer

Wild Ducks is building a real-time, event-driven forecasting platform with complex multi-tenant datasets, high-throughput ingestion, and strict SLA and correctness requirements. We’re looking for an engineer to own our data architecture end-to-end: modeling, schema evolution, performance, partitioning, RLS, observability, and alignment with our event system and forecasting engine.This role sits at the intersection of database engineering, data modeling, platform reliability, and product architecture. If you’re motivated by building durable data systems that scale, stay clean over time, and power critical forecasting logic, you’ll thrive here.What You’ll OwnDatabase Architecture & ModelingOwn the structure, evolution, and quality of our Postgres schemas across all domains (forecasting, demand ingestion, eventing, TSL, geolocation, etc.).Design high-integrity, multi-tenant-safe data models using UUID PKs, real foreign keys, RLS, partitioning, and row-level auditability.Work with engineering leadership to shape the canonical data model behind demand, installed base, forecasting, TSL, and analytics workloads.Ensure schemas align with business invariants and support future growth without fragmentation.Performance, Reliability & PartitioningDesign and maintain table partitioning strategies (time-based and tenant-based) for high-throughput workloads.Diagnose and resolve performance bottlenecks: query tuning, index planning, materialized views, caching layers.Build and maintain observability around database health, query performance, storage growth, and index efficiency.Data Engineering & PipelinesPartner with ingestion, forecasting, and orchestration teams to ensure all pipelines read/write safely with proper RLS, locking discipline, and idempotency.Improve and harden cross-service data flows (Django ORM, SQLAlchemy, Temporal activities, Pulsar event persistence).Maintain the event → DB → projection/read-model lifecycle and ensure correctness across domains.Governance, Integrity & ToolingDefine and enforce our schema migration standards (zero-downtime migrations, NOT VALID → VALID foreign keys, safe column evolution).Build internal tools that keep the data platform clean (schema diffing, RLS linters, migration validation pipelines).Partner with the CTO to establish data governance, retention, backup, and archiving policies.Strategic & Cross-Functional WorkWork closely with product and engineering teams to design data models that match domain needs (parts, stock_location, routing, forecasting, TSL).Support analytics and insights teams by defining readable, durable read models.Participate in architectural reviews and major feature design to ensure the data layer is scalable and future-proof.Who You AreCore TraitsYou think in systems, invariants, and lifecycle, not just tables.You are obsessive about correctness, consistency, and clarity in data structures.You communicate clearly and collaboratively with engineers across backend, eventing, forecasting, and DevOps.You enjoy both designing new models and untangling old ones.Experience That HelpsDeep familiarity with Postgres: querying, indexing, performance tuning, partitioning, RLS, WAL, explain plans.Experience designing multi-tenant schemas with strong security boundaries.Knowledge of event-driven systems and how data behaves around Pulsar/Kafka, outbox patterns, and projection models.Comfort with Python (Django, SQLAlchemy) or equivalent backend frameworks.Experience building or evolving data architectures for SaaS, logistics, or forecasting platforms.Experience supporting high-throughput ingestion systems.Experience designing data warehouse adjacencies (future ClickHouse, OLAP layers) is a plus.Bonus: familiarity with forecasting concepts (demand, installed base, TSL, planning nodes) or willingness to learn.What You’ll Work WithPostgres 14+ with multi-tenant RLSDjango ORM, custom SQL, Temporal workflowsPulsar event streams, outbox-driven persistenceRedis, caching layers, materialized viewsGKE Autopilot, Terraform-managed infraOpenTelemetry, query metrics, database monitoringTools for schema evolution (Sqitch, Alembic-like tooling, Django migrations)How We WorkHybrid: remote-friendly with periodic in-person design sessions.No silos: you work across forecasting, ingestion, eventing, and infra.High trust, high ownership: your work defines the foundation of our system for years ahead.Tight collaboration with platform, data science, and product teams.