Founding Engineer — Core Systems

Imprezia is building the monetization infrastructure for the AI era.

AI agents are becoming the primary interface to the internet. Every interface that captures attention eventually requires a monetization layer. That outcome is inevitable. What does not yet exist is the infrastructure that makes monetization inside AI-native experiences work at scale. We are building it.

Our system operates in real time, is intent-aware, and is designed for how people actually interact with AI agents. The team comes from MIT, Meta AI Ads, Amazon Ads, and more. We have built and operated systems that move billions of dollars per day. We know where legacy ad tech breaks. This company exists to replace those assumptions.

This is a founding engineering role with real ownership. You will design and build the core decisioning, retrieval, ranking, and delivery systems that determine what appears inside AI-generated content, when it appears, and why. You will make architectural decisions early, ship under uncertainty, and operate systems under real demand and revenue pressure.

We are building primitives of monetization for the next generation of Internet.

Apply if you want to help define how AI systems fund themselves before the rules are written.

What you’ll do

• Own end-to-end design and operation of real-time decisioning systems under strict latency and scale constraints
• Build and evolve low-latency retrieval and ranking infrastructure from zero to production
• Make and live with architectural tradeoffs around throughput, correctness, and failure modes

What we’re looking for

• Exceptional systems engineer with experience building distributed infrastructure at scale
• Strong intuition for latency-sensitive, high-throughput services
• Ownership mindset; ad tech or marketplace experience is a plus, not a requirement

Why this role is different

• You own core primitives, not edge features
• Your work directly shapes revenue and product direction
• You help define AI monetization infrastructure before it becomes standardized