Director, Data Products & Platform Strategy

Stream Companies•West Chester, PA

About The Position

This is a senior role built around two connected purposes that share a common thread: making messy, multi-source data useful, trustworthy, and commercially sharp. First, you'll be the data product owner and principal analyst for Ask Liv, the AI assistant we're building on Snowflake Cortex inside OrangeOS, set to launch later this year. You'll be the subject-matter expert on what our data actually is, what it can support, and how it should show up — accurately and reliably — in a chat response. Second, you'll be the strategic and technical lead driving an offshore development team building Liquid Shopper, our new multi-tenant dealer-website platform. That means owning the data integration strategy and architecture the platform is built on — including diving into dealership back-office systems to understand what they hold and how it creates value — and bringing the retail marketing judgment needed to turn dealer and merchandising requirements into a build the offshore team can execute well. You'll report directly to the VP of Product and work closely with product, engineering, and data teams across both initiatives.

Requirements

Extensive experience architecting and integrating data solutions across large, multi-source environments (CRM, transactional, inventory/product feeds, marketing/campaign data). Deep tenure on any single warehouse platform matters less than the ability to reason about data architecture broadly — plus the aptitude to quickly become a Snowflake power user, accelerated by AI copilots such as Snowflake's Cortex/CoCo-style assistants.
Hands-on experience building real applications or workflows using AI coding agents and assistants (e.g., Claude Code, GitHub Copilot, Cursor, or similar) — comfortable directing and reviewing AI-generated work, not only writing code by hand. This is central to how we intend to build Liquid Shopper.
Working familiarity with GenAI or LLM-powered products — RAG concepts (chunking, retrieval, embeddings) and output-quality evaluation.
A track record of owning data integration across multiple sources for a customer-facing product — comfortable with both the pipes and what flows through them.
Experience translating business or marketing requirements into technical specifications and acceptance criteria for a development team — including distributed or offshore teams — without relying on formal management authority.
A retail marketing or merchandising background: familiarity with promotions, incentive or offer structures, personalization, or customer-facing digital retail experiences. Automotive experience is a strong plus, not a requirement.
Strong data manipulation, automation, and light AI/workflow development skills.
Excellent communication across audiences — equally comfortable briefing an executive on a product trade-off and a developer on a data model.
Genuine comfort driving strategy, architecture, and delivery quality directly — this role is measured on impact, not on the size of a team underneath you.

Nice To Haves

Direct Snowflake experience — semantic/metric layer design, RBAC and access governance, or Snowflake Cortex (Analyst, Search, or Agents). Valuable, but not required if you can ramp quickly.
Experience standing up or scaling a multi-tenant SaaS platform.
Prior experience directing offshore or outsourced engineering teams.
Automotive retail, dealer marketing, or martech background.
Familiarity with PHP, Laravel, or Vue.js (Stream's core stack), or comparable frameworks.

Responsibilities

Own the data and AI foundation behind Ask Liv (OrangeOS)
Semantic layer & governance — Own the semantic views and business definitions that encode what counts as a sale, a lead, or a qualifying offer, and administer Snowflake access controls for the AI platform. Audit and refactor existing data models as needed.
Retrieval & output quality — Control retrieval accuracy through chunk sizing, document parsing, and embedding configuration; reconcile Ask Liv's outputs against reporting models; audit vehicle offer and payment calculations for compliance risk; and expand evaluation test sets from real output failures.
Pipeline ownership — Maintain and monitor ingestion pipelines feeding the platform — inventory, dealer feeds, CRM, and marketing data — and alert on freshness or reliability issues before they reach the chatbot.
Cross-functional translation — Act as the liaison between product, engineering, and data — converting business questions into OrangeOS data structures and defining acceptance criteria for data-related work.
Lead the data and delivery strategy for Liquid Shopper
Back-office systems integration — Dive into dealership back-office systems — the DMS (dealer management system), inventory management system (IMS), and CRM — to understand what data they hold and map it to real functionality and value inside Liquid Shopper.
Data integration & architecture — Own the data integration strategy underpinning a multi-tenant platform — bringing together dealer inventory, incentive and pricing data, and marketing content across many client sites without losing tenant-level nuance.
Retail marketing translation — Turn dealer and retail-marketing requirements — offers, promotions, personalization, merchandising — into technical specifications and acceptance criteria the offshore team can build against.
AI-first delivery — Champion an AI-assisted, agentic development approach with the offshore team — using AI coding agents to increase velocity and quality and reduce the software engineering headcount the platform requires as it scales.
Offshore build coordination — Drive sprint planning, backlog priority, and quality review with the offshore development team as the senior technical voice on the project.
Roadmap partnership — Serve as a strategic partner to the VP of Product on scope, sequencing, and trade-offs as Liquid Shopper moves from build to a sellable, multi-tenant product.