Principal Engineer, Platform & Agents

Principal Engineer, Platform & Agents We're building streaming tools for the top creators in the space The streaming space is one of the most creative, fastest-moving corners of the internet, and the tools serving creators have not kept up. We think creators deserve a platform that feels like a collaborator instead of a control panel: something that meets them where they are, scales with their channel, and helps them build community without an engineering degree. That's the bet we're making, and we're hiring a Principal Engineer to co-lead the technical direction of this product alongside our existing engineering architect. The role This is a deliberately two-sided role, and we need someone who can hold both sides at once. On the platform side, you'll own the architecture that ingests events from every major streaming service and fans them out reliably to creators and their audiences. The hard part isn't writing the first integration. The hard part is keeping a long-lived connection alive for hours under real load, surviving OAuth refreshes and silent disconnects, deduping events that arrive twice, and then designing the system so the next platform integration takes days instead of weeks. Reliability is the product. When a creator goes live and our connection drops, they notice immediately, and so do their viewers. On the agent side, you'll help shape and ship the AI experiences that make our platform feel different. Tool calling, context engineering, model routing, cost and latency guardrails at production scale. We care about agents that do real work for creators, not chatbots that demo well. If you've only done one of these well, you'll struggle here. The platform without the agent is a commodity. The agent without the platform is a demo. We need both, in one head. What you'll do Co-own the architecture of our realtime platform, including event ingestion, normalization across inconsistent third-party APIs, and the systems that fan events out to widgets and automations. Design our integration layer so adding the next streaming platform is a small, well-paved path instead of a months-long project. Set the technical direction for our move from a Node and React monolith toward selectively extracted Go services where throughput and cost demand it. Help shape and build our agent experiences end to end, from the tool surface and orchestration layer to the UX patterns that make it feel like a collaborator.

Own reliability for the platform:

connection lifecycle, OAuth, observability, on-call posture, and the boring fundamentals that keep creators online. Mentor a small engineering team and help us hire the next two engineers. What we need from you These are non-negotiable. If you're missing one, this role is not for you yet. You've shipped a production agent that does real work. Not a prototype, not a hackathon. Something with users, cost constraints, and failure modes you've debugged at 2am. You understand tool calling, context engineering, and where current models break. You've operated a system handling high-volume realtime events. WebSockets at scale, long-lived connections that survive OAuth refreshes and silent disconnects, backpressure, dedup, idempotency. You know why holding a connection to a third-party API is hard, and you've done it. You can design an integration layer that scales by team velocity, not just by traffic. You've built systems where adding the next vendor, partner, or platform is a clean, paved path. You know what a good adapter pattern looks like and what a bad one feels like six months later. You're fluent in TypeScript and Go , and you reach for the right one based on the problem instead of preference. You've worked with FFmpeg in production. Generating, transcoding, or piping media programmatically. You know its sharp edges. You have real DevOps depth. Kubernetes, CI/CD, IAM, observability. You can't be allergic to YAML. You understand creators. You've either built tools for creators, worked at a creator-economy or live-video company, or you create content yourself. We'll filter hard on this. Taste for the medium is something we cannot teach you. Bonuses we'll notice You've built or operated an MCP server in production. You've integrated with major streaming platform APIs and can tell us which one made you want to quit engineering. You have a track record of being a second technical hire and not destroying the first one's work. What this isn't A pure research role. We ship. An IC role with no leadership weight. You'll co-set technical direction with our existing technical lead. A safe role. We're a small team making a big bet, and the failure modes are real. About Blerp Founded out of a dorm room at the University of Utah, originally a Shia LaBeouf soundboard that hit 100K downloads. Ten years later we're a creator-tools platform across every major streaming service, with real revenue, real customers, and a deliberate subscription pivot underway. Techstars and Amazon-backed. Small team, long memory, no layers. If this sounds like your kind of space, come join us!

Pay:

$75,000.00 - $150,000.00 per year

Benefits:

Flexible schedule Relocation assistance

Work Location:

Hybrid remote in Salt Lake City, UT