Staff Data Engineer

Harbor Compliance is building its first end-to-end data platform — unifying fragmented data from HubSpot, financial systems, and HRIS into a single, AI-augmented source of truth for executive decision-making. As Staff Data Engineer, you'll be a foundational technical hire, working closely with the Sr. Director of Data Platform & Analytics to design and build the real-time, AI-ready data infrastructure that powers this platform. This is a hands-on role for someone who wants to build from the ground up, not maintain what already exists.

Key Responsibilities

• Design, build, and own near real-time data pipelines (CDC, streaming ingestion, event-driven architectures) as the backbone of the platform's data flow
• Evaluate, implement, and maintain vector database infrastructure and embedding pipelines to support AI-augmented use cases (semantic search, retrieval-augmented generation, AI agents acting on company data).
• Design and build ELT/ETL pipelines ingesting data from our platform, financial platforms, CRM (HubSpot) and HRIS, feeding both real-time and batch use cases.
• Partner with the Sr. Director to architect the underlying warehouse/lakehouse as a supporting system of record — the storage layer beneath the streaming and AI infrastructure.
• Build lightweight transformation layers (e.g., dbt) as needed to enable our Analytics Engineering team translate raw data into business-ready datasets aligned to core metrics like ARR, CAC, and churn.
• Own pipeline reliability and observability — monitoring, automated failure alerting, and lineage tracking across both streaming and batch pipelines.
• Build the technical foundation for self-service and AI-powered reporting, partnering with BI, Product & Engineering stakeholders on recurring executive and departmental reports.
• Implement data governance practices, including documentation standards and access controls.
• Partner cross-functionally with Finance, Marketing, Customer Success, and Operations to translate data needs into reliable, low-latency data products.
• Leverage AI-augmented development workflows (e.g., Claude Code) to accelerate pipeline development and documentation.

Requirements

• 7+ years of hands-on data engineering experience, with meaningful depth in streaming/event-driven systems, not just batch pipelines.
• Proven experience designing and building near real-time pipelines from scratch (e.g., Kafka, Kinesis, Flink, Debezium/CDC) in a production environment.
• Hands-on production experience with vector databases and embeddings (e.g., Zilliz, Pinecone, Weaviate, pgvector, Milvus) — ideally having built this infrastructure from the ground up rather than just consuming a managed AI feature.
• Advanced proficiency in SQL and Python.
• Working knowledge of a cloud warehouse/lakehouse platform (Snowflake, BigQuery, or Databricks) and dbt — you'll use these, but they're the storage/transform layer supporting the streaming and AI work, not the main focus.