We're building the first AI-native performance marketing platform for Streaming TV. DTC and e-commerce brands use Upscale to turn their best brand, product, and social content into high-performing CTV campaigns — with AI-driven creative optimization, advanced targeting, and digital measurement that drives real outcomes (CPA, site visits, purchases) and measurable lift across search, social, and other digital channels.

We're seed-stage, funded, and moving fast.

The Role

You'll own the backend infrastructure that powers our AI and advertising systems, with a focus on three core areas:

• Process & orchestration — Our "Control Room" coordinates multiple LLMs and generative AI models for video creative optimization. You'll design the orchestration layer, manage complex agentic workflows, and build the reliability infrastructure around them. This is the heart of the role.
• System monitoring & testing — You'll build the observability, alerting, and testing infrastructure that keeps our systems healthy. Think structured logging, distributed tracing, health checks, integration test suites, and the kind of production confidence that lets a small team move fast without breaking things.
• Data pipelines — You'll build and maintain the pipelines that move and transform data across our platform — ingesting campaign and creative performance data, feeding downstream systems, and ensuring data arrives reliably and on time.

These systems feed into a real-time bidding pipeline, so you'll need to understand RTB concepts and latency constraints, but you won't be building the bidding infrastructure from scratch.

You'll work primarily in Python (FastAPI, async patterns), PostgreSQL, and Kubernetes, deploying on cloud infrastructure. This is a small team where you'll have significant architectural influence.

What we're looking for

• 5+ years building backend systems in Python, with genuine depth — not just framework experience but understanding of concurrency, performance profiling, and system design tradeoffs.
• Hands-on experience with at least one of: agentic AI systems, complex workflow orchestration, or production observability at scale.
• Comfort with Kubernetes, Helm, and infrastructure-as-code. You've written Helm charts, set up monitoring, and debugged production issues at 2 AM.
• Experience building CI/CD pipelines, test automation frameworks, or production monitoring stacks.
• Strong data pipeline experience — you've built systems that move and transform large volumes of data reliably.
• Distributed systems intuition. You think about failure modes, consistency tradeoffs, and observability as first-class concerns.
• Clear communication. We're a small, remote-friendly team across multiple time zones. Writing well and thinking out loud matters.

Nice to have

• Experience integrating LLMs or generative AI models into production systems.
• Background in event-driven architectures (Redis Streams, Kafka, or similar).
• Adtech familiarity — you don't need deep RTB experience, but understanding the domain accelerates everything.
• Prior startup experience where you wore multiple hats.

What you get

• Real equity in a seed-stage company with a large addressable market — this isn't a token grant.
• Direct architectural influence over AI-native systems that are genuinely novel, not another CRUD app with an LLM bolted on.
• A founding engineering team led by a CTO with 25+ years in distributed systems, including Principal Architect at Salesforce and multiple successful exits.
• Hybrid flexibility (2–3 days/week in our San Mateo or New York office).
• Competitive salary, benefits, and the usual startup perks — but mostly, genuinely hard problems worth solving.

Upscale AI is an equal opportunity employer.