Director, Data Engineering - Data & AI Platform - Job ID 959

Kapitus is a small business financing company building a next-generation enterprise data and AI capability. The Director of Data Engineering will serve as the senior technical leader for the modernization program, owning the architecture, engineering standards, platform delivery, reliability, and distributed engineering teams for a Snowflake-centered data platform.

Responsibilities:

• Own the engineering and solution architecture of the data platform within approved enterprise and data architecture standards including layered warehouse design, data modeling, ingestion and transformation patterns, orchestration, environment strategy, and promotion controls and keep delivery conformant to it
• Translate approved data architecture, MDM, ontology, semantic, and data-contract standards into enforceable engineering patterns so identity, meaning, metrics, and relationships remain consistent across products
• Set and enforce engineering standards: repository structure, branching and CI/CD, code review, testing and data quality checks, naming and documentation, and performance and cost discipline
• Lead design reviews and architecture decision forums; document decisions, manage exceptions with owners and expiry dates, and prevent conformance debt from silently accumulating
• Design for reuse: shared ingestion frameworks, certified transformation patterns, common serving structures, and semantic consistency — so each delivery wave gets faster, not just bigger
• Stay technically hands-on: review critical code and designs, prototype or intervene on the highest-risk components when needed, tune warehouse performance and cost, and establish the engineering bar through direct technical engagement
• Lead the engineering build of governed business data products end to end: source onboarding, ELT pipeline development, data modeling, quality controls, serving and consumption structures, and production release
• Direct the migration and decommissioning of legacy analytics workflows onto the modern platform. Reuse, refactor, retire, or rebuild decisions grounded in analysis, not habit
• Own platform hardening and readiness: environments, security and access patterns, orchestration reliability, and operational readiness the platform is a product, not a project by-product
• Own engineering reliability for production data services including observability, service-level objectives, incident and problem management, recovery patterns, runbooks, and resilience/continuity requirements with recurring failures driven to root-cause closure
• Engineer for the AI/ML workstream: feature-ready data, ML pipeline integration, and the data foundations that GenAI and agentic workloads depend on synchronized with platform readiness rather than bolted on
• Deliver against the program’s milestone gates: vendor verification, business/customer validation, production acceptance. Code complete is not product complete your teams’ work is done when it is validated, evidenced, and accepted
• Lead and develop a blended engineering organization: internal engineers, partner delivery pods, and offshore teams across time zones
• Make the onshore/offshore model actually work: clean handoffs, clear design specifications before build starts, overlap windows used deliberately, and quality standards that do not vary by location
• Uphold separation of duties between build and validation teams, and give validators what they need to verify work independently
• Assess partner engineering quality directlyreview their designs and code, challenge estimates from technical knowledge, and hold delivery partners to the same standards as internal teams
• Hire, coach, and grow engineers; set clear expectations; and build a culture where problems surface early and evidence beats assertion
• Partner with the Technical Program Manager on sequencing, capacity, dependency management, and gate readiness you own technical feasibility and quality; the TPM owns the integrated plan and commercial controls
• Work with data governance so controls ship with the product: quality rules, lineage, classification handling, and catalog readiness built into pipelines rather than retrofitted
• Engage business owners and architecture directly — explain technical trade-offs in business terms and drive timely technical decisions with accountable owners

Requirements:

• 10+ years of data engineering experience
• 5+ years leading data engineering teams through enterprise data platform builds or modernizations
• Deep hands-on Snowflake expertise with warehouse and database design, performance tuning, cost and workload optimization, security and access patterns (roles, masking, row-level security), data sharing, and operational administration at production scale
• Hands-on dbt mastery including project architecture, layered modeling conventions, testing and documentation, macros and packages, CI/CD integration, and running dbt across multiple teams and environments while preventing uncontrolled project, model, and dependency sprawl
• Broad modern data stack fluency with orchestration (e.g., Airflow, Dagster, or similar), ingestion and CDC tooling, streaming and batch patterns, data catalogs and lineage, semantic/BI layers, and legacy workflow migration (e.g., Alteryx, SSIS, or similar)
• Strong software engineering fundamentals must be an expert in SQL and have solid experience with Python, Git-based workflows, CI/CD, automated testing, and infrastructure-as-code awareness; you hold data engineering to software engineering standards
• Architecture credibility you have designed platforms, not just built on them: data modeling at enterprise scale, layered architectures, master

data and semantic consistency concerns, and the judgment to know when a pattern is worth standardizing

• Proven distributed team leadership experience leading onshore/offshore engineering teams, including partner and vendor pods, with quality and accountability consistent across locations
• AI/ML platform literacy you must understand what ML and GenAI workloads demand of a data platform (feature pipelines, training and inference data flows, retrieval foundations, monitoring) well enough to engineer for them and to lead teams that do
• Strong communication skills you can defend an architecture to engineers, explain a trade-off to executives, and write design documentation people actually use
• Databricks experience is a plus
• Consulting or professional services delivery background preferred
• Experience in financial services in lending, banking, fintech, or another regulated environment including working under data governance, model risk, and audit expectations
• SnowPro certifications (Core, Advanced Architect, or Data Engineer), dbt certification, or cloud (AWS/Azure) certifications valued as evidence of rigor, not as a substitute for demonstrated delivery experience
• Experience with data product operating models (data mesh / data product thinking, data contracts, certified metrics) and with building platforms that serve them
• FinOps experience measuring and managing cloud data platform cost as an engineering discipline
• Experience standing up engineering practices from scratch on a program: standards, CI/CD, review culture, validation streams, and operational readiness