The Opportunity:

ZoomInfo built the industry's most sophisticated GTM data acquisition infrastructure. Now we're applying that same rigor to context engineering—the emerging discipline that determines whether AI systems deliver transformative value or incremental improvement.

This role architects the context layer powering our AI intelligence across Copilot, GTM Studio, and MarketingOS. You'll transform how ZoomInfo's agentic workflows access, compress, and deliver precisely the right information at exactly the right moment. The impact is organization-wide: every AI interaction, every intelligent recommendation, every autonomous agent action depends on the context infrastructure you’ll build.

We've transitioned to AI-first product thinking company-wide. The context pipelines exist but remain nascent—creating a rare opportunity to define architectural patterns and platform standards that compound value across multiple product teams in the years to come.

What You'll Do:

Architect Context Acquisition Pipelines Design and optimize how ZoomInfo retrieves, transforms, and delivers context from our semantic data layer, memory systems, and data producers. You'll balance retrieval quality against latency and cost constraints, implementing hybrid search strategies, intelligent caching, and context compression techniques that maintain information density while respecting token budgets.

Own the Context Layer Platform Build infrastructure serving multiple product teams—Copilot, GTM Studio, MarketingOS—as internal customers. Establish API contracts, developer experience standards, and integration patterns that accelerate feature velocity. Maintain the delicate balance between providing flexible building blocks and opinionated solutions that encode best practices.

Drive Quality Through Measurement Implement evaluation frameworks using RAGAS metrics and custom benchmarks. Monitor retrieval precision, context relevance, hallucination rates, and system performance in production. Translate quality signals into architectural improvements, working closely with ML engineers to iterate on embedding models, reranking strategies, and retrieval algorithms.

Navigate Emerging Research Context engineering evolves weekly. You'll continuously evaluate innovations—GraphRAG for multi-hop reasoning, test-time compute scaling, multimodal retrieval, compression techniques—determining which advances warrant production investment versus which remain academic curiosities. Bring external best practices to ZoomInfo while contributing learnings back to the broader community.

Orchestrate Cross-Functional Execution Translate between three distinct worlds: ML engineers optimizing retrieval algorithms, platform engineers building scalable infrastructure, and product teams shipping customer features. Establish communication cadences, prioritization frameworks, and decision-making processes that balance urgent requests against strategic platform development.

What You’ll Bring:

• 6-8 years of product management experience with 2+ years in ML/AI infrastructure
• Direct experience with production RAG systems, vector databases, or semantic search, context management
• Experience with graph databases (e.g. Neo4j) 
• Track record building platform products serving multiple internal or external customers
• Familiarity with context compression, embedding models, and retrieval evaluation frameworks
• History of defining product vision in nascent technical domains where best practices are still emerging

Who You Are:

Technical Foundation Expert-level understanding of RAG system architecture—you can discuss embedding dimensionality trade-offs, vector database indexing strategies, and reranking approaches with depth. You've built or significantly contributed to production retrieval systems, not just managed them at arm's length. Python and SQL proficiency enables you to review code, analyze retrieval issues, and prototype solutions for concept validation.

Platform Product Mindset Experience building infrastructure products where internal engineering teams are your customers. You measure success through downstream product velocity improvements and developer satisfaction scores, not just uptime metrics. You understand platform economics—how each additional team using your infrastructure increases its value through shared learnings and amortized costs.

Intellectual Velocity You read recent research papers from arXiv, ACL, NeurIPS. You prototype emerging techniques to understand their practical constraints. You maintain strong opinions weakly held, updating your architectural assumptions as evidence accumulates. The discipline moves too fast for static expertise—continuous learning is non-negotiable.

Strategic Communication You translate between technical depth and business impact fluently. You can explain to executives why implementing GraphRAG takes 6 months but unlocks $10M in product capabilities. You can communicate to engineers why business constraints require shipping "good enough" in 3 weeks rather than "optimal" in 3 months. You influence without formal authority through data, clear reasoning, and earned credibility.

The Environment:

Reporting & Collaboration Report to the Senior Product Director for Context Engineering, Semantic Data Layer, and Agentic Memory within ZoomInfo's Intelligence team. Work alongside PMs responsible for signals and ML scoring/recommendation models. Together, you ensure our agentic workflows fill context windows with high-quality, information-dense content exactly when needed.

Pace & Problems Fast-moving engineering team that understands the space. Company-wide AI adoption push creates both urgency and opportunity. Expect interesting problems: How do we maintain sub-200ms retrieval latency at scale? When does GraphRAG justify its indexing cost? How do we balance context freshness with cache efficiency? You'll shape answers that become architectural patterns across the organization.

Impact Define a nascent discipline at a company that's already AI-first in product thinking and organizational structure. Your architectural decisions compound—every improvement to context quality multiplies across Copilot, GTM Studio, MarketingOS, and future products we haven't imagined yet. This is infrastructure work with direct line-of-sight to customer value.

#LI-VC1

#LI-Remote